Merge pull request #97 from justadudewhohacks/train-landmark-models

README.md

@@ -43,7 +43,8 @@ Table of Contents:
 
 ### Face Landmarks
 
-![preview_face_landmarks_boxes](https://user-images.githubusercontent.com/31125521/41507933-65f9b642-723c-11e8-8f4e-aab13303e7ff.jpg)
+![face_landmarks_boxes_1](https://user-images.githubusercontent.com/31125521/46063403-fff9f480-c16c-11e8-900f-e4b7a3828d1d.jpg)
+![face_landmarks_boxes_2](https://user-images.githubusercontent.com/31125521/46063404-00928b00-c16d-11e8-8f29-e9c50afd2bc8.jpg)
 
 ![preview_face_landmarks](https://user-images.githubusercontent.com/31125521/41507950-e121b05e-723c-11e8-89f2-d8f9348a8e86.png)
 
@@ -57,10 +58,6 @@ Table of Contents:
 
 ![mtcnn-preview](https://user-images.githubusercontent.com/31125521/42756818-0a41edaa-88fe-11e8-9033-8cd141b0fa09.gif)
 
-### Face Alignment
-
-![preview_face_alignment](https://user-images.githubusercontent.com/31125521/41526994-1a690818-72e6-11e8-8f3c-d2cf31fe517b.jpg)
-
 <a name="running-the-examples"></a>
 
 ## Running the Examples
@@ -89,7 +86,7 @@ The face detection model has been trained on the [WIDERFACE dataset](http://mmla
 
 ### Face Detection - Tiny Yolo v2
 
-The Tiny Yolo v2 implementation is a very performant face detector, which can easily adapt to different input image sizes, thus can be used as an alternative to SSD Mobilenet v1 to trade off accuracy for performance (inference time). In general the models ability to locate smaller face bounding boxes is not as accurate as SSD Mobilenet v1. 
+The Tiny Yolo v2 implementation is a very performant face detector, which can easily adapt to different input image sizes, thus can be used as an alternative to SSD Mobilenet v1 to trade off accuracy for performance (inference time). In general the models ability to locate smaller face bounding boxes is not as accurate as SSD Mobilenet v1.
 
 The face detector has been trained on a custom dataset of ~10K images labeled with bounding boxes and uses depthwise separable convolutions instead of regular convolutions, which ensures very fast inference and allows to have a quantized model size of only 1.7MB making the model extremely mobile and web friendly. Thus, the Tiny Yolo v2 face detector should be your GO-TO face detector on mobile devices.
 
@@ -113,9 +110,7 @@ The neural net is equivalent to the **FaceRecognizerNet** used in [face-recognit
 
 ### 68 Point Face Landmark Detection
 
-This package implements a CNN to detect the 68 point face landmarks for a given face image.
-
-The model has been trained on a variety of public datasets and the model weights are provided by [yinguobing](https://github.com/yinguobing) in [this](https://github.com/yinguobing/head-pose-estimation) repo.
+This package implements a very lightweight and fast, yet accurate 68 point face landmark detector. The default model has a size of only 350kb and the tiny model is only 80kb. Both models employ the ideas of depthwise separable convolutions as well as densely connected blocks. The models have been trained on a dataset of ~35k face images labeled with 68 face landmark points.
 
 <a name="usage"></a>
 
@@ -145,6 +140,7 @@ Assuming the models reside in **public/models**:
 await faceapi.loadFaceDetectionModel('/models')
 // accordingly for the other models:
 // await faceapi.loadFaceLandmarkModel('/models')
+// await faceapi.loadFaceLandmarkTinyModel('/models')
 // await faceapi.loadFaceRecognitionModel('/models')
 // await faceapi.loadMtcnnModel('/models')
 // await faceapi.loadTinyYolov2Model('/models')
@@ -155,19 +151,18 @@ As an alternative, you can also create instance of the neural nets:
 ``` javascript
 const net = new faceapi.FaceDetectionNet()
 // accordingly for the other models:
-// const net = new faceapi.FaceLandmarkNet()
+// const net = new faceapi.FaceLandmark68Net()
+// const net = new faceapi.FaceLandmark68TinyNet()
 // const net = new faceapi.FaceRecognitionNet()
 // const net = new faceapi.Mtcnn()
 // const net = new faceapi.TinyYolov2()
 
 await net.load('/models/face_detection_model-weights_manifest.json')
 // await net.load('/models/face_landmark_68_model-weights_manifest.json')
+// await net.load('/models/face_landmark_68_tiny_model-weights_manifest.json')
 // await net.load('/models/face_recognition_model-weights_manifest.json')
 // await net.load('/models/mtcnn_model-weights_manifest.json')
 // await net.load('/models/tiny_yolov2_separable_conv_model-weights_manifest.json')
-
-// or simply load all models
-await net.load('/models')
 ```
 
 Using instances, you can also load the weights as a Float32Array (in case you want to use the uncompressed models):

examples/public/styles.css

@@ -67,6 +67,12 @@
   left: 0;
 }
 
+.overlay {
+  position: absolute;
+  top: 0;
+  left: 0;
+}
+
 #facesContainer canvas {
   margin: 10px;
 }
\ No newline at end of file

src/faceLandmarkNet/FaceLandmark68Net.ts

 import * as tf from '@tensorflow/tfjs-core';
-import { NetInput } from 'tfjs-image-recognition-base';
-import { convLayer, ConvParams } from 'tfjs-tiny-yolov2';
+import { NetInput, normalize } from 'tfjs-image-recognition-base';
+import { ConvParams } from 'tfjs-tiny-yolov2';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
 
+import { depthwiseSeparableConv } from './depthwiseSeparableConv';
 import { extractParams } from './extractParams';
 import { FaceLandmark68NetBase } from './FaceLandmark68NetBase';
 import { fullyConnectedLayer } from './fullyConnectedLayer';
 import { loadQuantizedParams } from './loadQuantizedParams';
-import { NetParams } from './types';
+import { DenseBlock4Params, NetParams } from './types';
 
-function conv(x: tf.Tensor4D, params: ConvParams): tf.Tensor4D {
-  return convLayer(x, params, 'valid', true)
-}
+function denseBlock(
+  x: tf.Tensor4D,
+  denseBlockParams: DenseBlock4Params,
+  isFirstLayer: boolean = false
+): tf.Tensor4D {
+  return tf.tidy(() => {
+    const out1 = tf.relu(
+      isFirstLayer
+        ? tf.add(
+          tf.conv2d(x, (denseBlockParams.conv0 as ConvParams).filters, [2, 2], 'same'),
+          denseBlockParams.conv0.bias
+        )
+        : depthwiseSeparableConv(x, denseBlockParams.conv0 as SeparableConvParams, [2, 2])
+    ) as tf.Tensor4D
+    const out2 = depthwiseSeparableConv(out1, denseBlockParams.conv1, [1, 1])
+
+    const in3 = tf.relu(tf.add(out1, out2)) as tf.Tensor4D
+    const out3 = depthwiseSeparableConv(in3, denseBlockParams.conv2, [1, 1])
+
+    const in4 = tf.relu(tf.add(out1, tf.add(out2, out3))) as tf.Tensor4D
+    const out4 = depthwiseSeparableConv(in4, denseBlockParams.conv3, [1, 1])
 
-function maxPool(x: tf.Tensor4D, strides: [number, number] = [2, 2]): tf.Tensor4D {
-  return tf.maxPool(x, [2, 2], strides, 'valid')
+    return tf.relu(tf.add(out1, tf.add(out2, tf.add(out3, out4)))) as tf.Tensor4D
+  })
 }
 
 export class FaceLandmark68Net extends FaceLandmark68NetBase<NetParams> {
 
   constructor() {
-    super('FaceLandmark68Net')
+    super('FaceLandmark68LargeNet')
   }
 
   public runNet(input: NetInput): tf.Tensor2D {
@@ -27,27 +47,21 @@ export class FaceLandmark68Net extends FaceLandmark68NetBase<NetParams> {
     const { params } = this
 
     if (!params) {
-      throw new Error('FaceLandmark68Net - load model before inference')
+      throw new Error('FaceLandmark68LargeNet - load model before inference')
     }
 
     return tf.tidy(() => {
-      const batchTensor = input.toBatchTensor(128, true).toFloat() as tf.Tensor4D
-
-      let out = conv(batchTensor, params.conv0)
-      out = maxPool(out)
-      out = conv(out, params.conv1)
-      out = conv(out, params.conv2)
-      out = maxPool(out)
-      out = conv(out, params.conv3)
-      out = conv(out, params.conv4)
-      out = maxPool(out)
-      out = conv(out, params.conv5)
-      out = conv(out, params.conv6)
-      out = maxPool(out, [1, 1])
-      out = conv(out, params.conv7)
-      const fc0 = tf.relu(fullyConnectedLayer(out.as2D(out.shape[0], -1), params.fc0))
-
-      return fullyConnectedLayer(fc0, params.fc1)
+      const batchTensor = input.toBatchTensor(112, true)
+      const meanRgb = [122.782, 117.001, 104.298]
+      const normalized = normalize(batchTensor, meanRgb).div(tf.scalar(255)) as tf.Tensor4D
+
+      let out = denseBlock(normalized, params.dense0, true)
+      out = denseBlock(out, params.dense1)
+      out = denseBlock(out, params.dense2)
+      out = denseBlock(out, params.dense3)
+      out = tf.avgPool(out, [7, 7], [2, 2], 'valid')
+
+      return fullyConnectedLayer(out.as2D(out.shape[0], -1), params.fc)
     })
   }
 
@@ -55,6 +69,7 @@ export class FaceLandmark68Net extends FaceLandmark68NetBase<NetParams> {
     return loadQuantizedParams(uri)
   }
 
+
   protected extractParams(weights: Float32Array) {
     return extractParams(weights)
   }

src/faceLandmarkNet/FaceLandmark68TinyNet.ts

+import * as tf from '@tensorflow/tfjs-core';
+import { NetInput, normalize } from 'tfjs-image-recognition-base';
+import { ConvParams } from 'tfjs-tiny-yolov2';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
+
+import { depthwiseSeparableConv } from './depthwiseSeparableConv';
+import { extractParamsTiny } from './extractParamsTiny';
+import { FaceLandmark68NetBase } from './FaceLandmark68NetBase';
+import { fullyConnectedLayer } from './fullyConnectedLayer';
+import { loadQuantizedParamsTiny } from './loadQuantizedParamsTiny';
+import { DenseBlock3Params, TinyNetParams } from './types';
+
+function denseBlock(
+  x: tf.Tensor4D,
+  denseBlockParams: DenseBlock3Params,
+  isFirstLayer: boolean = false
+): tf.Tensor4D {
+  return tf.tidy(() => {
+    const out1 = tf.relu(
+      isFirstLayer
+        ? tf.add(
+          tf.conv2d(x, (denseBlockParams.conv0 as ConvParams).filters, [2, 2], 'same'),
+          denseBlockParams.conv0.bias
+        )
+        : depthwiseSeparableConv(x, denseBlockParams.conv0 as SeparableConvParams, [2, 2])
+    ) as tf.Tensor4D
+    const out2 = depthwiseSeparableConv(out1, denseBlockParams.conv1, [1, 1])
+
+    const in3 = tf.relu(tf.add(out1, out2)) as tf.Tensor4D
+    const out3 = depthwiseSeparableConv(in3, denseBlockParams.conv2, [1, 1])
+
+    return tf.relu(tf.add(out1, tf.add(out2, out3))) as tf.Tensor4D
+  })
+}
+
+export class FaceLandmark68TinyNet extends FaceLandmark68NetBase<TinyNetParams> {
+
+  constructor() {
+    super('FaceLandmark68TinyNet')
+  }
+
+  public runNet(input: NetInput): tf.Tensor2D {
+
+    const { params } = this
+
+    if (!params) {
+      throw new Error('FaceLandmark68TinyNet - load model before inference')
+    }
+
+    return tf.tidy(() => {
+      const batchTensor = input.toBatchTensor(112, true)
+      const meanRgb = [122.782, 117.001, 104.298]
+      const normalized = normalize(batchTensor, meanRgb).div(tf.scalar(255)) as tf.Tensor4D
+
+      let out = denseBlock(normalized, params.dense0, true)
+      out = denseBlock(out, params.dense1)
+      out = denseBlock(out, params.dense2)
+      out = tf.avgPool(out, [14, 14], [2, 2], 'valid')
+
+      return fullyConnectedLayer(out.as2D(out.shape[0], -1), params.fc)
+    })
+  }
+
+  protected loadQuantizedParams(uri: string | undefined) {
+    return loadQuantizedParamsTiny(uri)
+  }
+
+  protected extractParams(weights: Float32Array) {
+    return extractParamsTiny(weights)
+  }
+}
\ No newline at end of file

src/faceLandmarkNet/depthwiseSeparableConv.ts

+import * as tf from '@tensorflow/tfjs-core';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
+
+export function depthwiseSeparableConv(
+  x: tf.Tensor4D,
+  params: SeparableConvParams,
+  stride: [number, number]
+): tf.Tensor4D {
+  return tf.tidy(() => {
+    let out = tf.separableConv2d(x, params.depthwise_filter, params.pointwise_filter, stride, 'same')
+    out = tf.add(out, params.bias)
+    return out
+  })
+}
\ No newline at end of file

src/faceLandmarkNet/extractParams.ts

 import { extractWeightsFactory, ParamMapping } from 'tfjs-image-recognition-base';
-import { extractConvParamsFactory, extractFCParamsFactory } from 'tfjs-tiny-yolov2';
 
+import { extractorsFactory } from './extractorsFactory';
 import { NetParams } from './types';
 
 export function extractParams(weights: Float32Array): { params: NetParams, paramMappings: ParamMapping[] } {
@@ -12,19 +12,16 @@ export function extractParams(weights: Float32Array): { params: NetParams, param
     getRemainingWeights
   } = extractWeightsFactory(weights)
 
-  const extractConvParams = extractConvParamsFactory(extractWeights, paramMappings)
-  const extractFCParams = extractFCParamsFactory(extractWeights, paramMappings)
+  const {
+    extractDenseBlock4Params,
+    extractFCParams
+  } = extractorsFactory(extractWeights, paramMappings)
 
-  const conv0 = extractConvParams(3, 32, 3, 'conv0')
-  const conv1 = extractConvParams(32, 64, 3, 'conv1')
-  const conv2 = extractConvParams(64, 64, 3, 'conv2')
-  const conv3 = extractConvParams(64, 64, 3, 'conv3')
-  const conv4 = extractConvParams(64, 64, 3, 'conv4')
-  const conv5 = extractConvParams(64, 128, 3, 'conv5')
-  const conv6 = extractConvParams(128, 128, 3, 'conv6')
-  const conv7 = extractConvParams(128, 256, 3, 'conv7')
-  const fc0 = extractFCParams(6400, 1024, 'fc0')
-  const fc1 = extractFCParams(1024, 136, 'fc1')
+  const dense0 = extractDenseBlock4Params(3, 32, 'dense0', true)
+  const dense1 = extractDenseBlock4Params(32, 64, 'dense1')
+  const dense2 = extractDenseBlock4Params(64, 128, 'dense2')
+  const dense3 = extractDenseBlock4Params(128, 256, 'dense3')
+  const fc = extractFCParams(256, 136, 'fc')
 
   if (getRemainingWeights().length !== 0) {
     throw new Error(`weights remaing after extract: ${getRemainingWeights().length}`)
@@ -32,17 +29,6 @@ export function extractParams(weights: Float32Array): { params: NetParams, param
 
   return {
     paramMappings,
-    params: {
-      conv0,
-      conv1,
-      conv2,
-      conv3,
-      conv4,
-      conv5,
-      conv6,
-      conv7,
-      fc0,
-      fc1
-    }
+    params: { dense0, dense1, dense2, dense3, fc }
   }
 }
\ No newline at end of file

src/faceLandmarkNet/extractParamsTiny.ts

+import { extractWeightsFactory, ParamMapping } from 'tfjs-image-recognition-base';
+
+import { extractorsFactory } from './extractorsFactory';
+import { TinyNetParams } from './types';
+
+export function extractParamsTiny(weights: Float32Array): { params: TinyNetParams, paramMappings: ParamMapping[] } {
+
+  const paramMappings: ParamMapping[] = []
+
+  const {
+    extractWeights,
+    getRemainingWeights
+  } = extractWeightsFactory(weights)
+
+  const {
+    extractDenseBlock3Params,
+    extractFCParams
+  } = extractorsFactory(extractWeights, paramMappings)
+
+  const dense0 = extractDenseBlock3Params(3, 32, 'dense0', true)
+  const dense1 = extractDenseBlock3Params(32, 64, 'dense1')
+  const dense2 = extractDenseBlock3Params(64, 128, 'dense2')
+  const fc = extractFCParams(128, 136, 'fc')
+
+  if (getRemainingWeights().length !== 0) {
+    throw new Error(`weights remaing after extract: ${getRemainingWeights().length}`)
+  }
+
+  return {
+    paramMappings,
+    params: { dense0, dense1, dense2, fc }
+  }
+}
\ No newline at end of file

src/faceLandmarkNet/extractorsFactory.ts

+import * as tf from '@tensorflow/tfjs-core';
+import { ExtractWeightsFunction, ParamMapping } from 'tfjs-image-recognition-base';
+import { extractConvParamsFactory, FCParams } from 'tfjs-tiny-yolov2';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
+
+import { DenseBlock3Params, DenseBlock4Params } from './types';
+
+export function extractorsFactory(extractWeights: ExtractWeightsFunction, paramMappings: ParamMapping[]) {
+
+  function extractSeparableConvParams(channelsIn: number, channelsOut: number, mappedPrefix: string): SeparableConvParams {
+    const depthwise_filter = tf.tensor4d(extractWeights(3 * 3 * channelsIn), [3, 3, channelsIn, 1])
+    const pointwise_filter = tf.tensor4d(extractWeights(channelsIn * channelsOut), [1, 1, channelsIn, channelsOut])
+    const bias = tf.tensor1d(extractWeights(channelsOut))
+
+    paramMappings.push(
+      { paramPath: `${mappedPrefix}/depthwise_filter` },
+      { paramPath: `${mappedPrefix}/pointwise_filter` },
+      { paramPath: `${mappedPrefix}/bias` }
+    )
+
+    return new SeparableConvParams(
+      depthwise_filter,
+      pointwise_filter,
+      bias
+    )
+  }
+
+  function extractFCParams(channelsIn: number, channelsOut: number, mappedPrefix: string): FCParams {
+    const weights = tf.tensor2d(extractWeights(channelsIn * channelsOut), [channelsIn, channelsOut])
+    const bias = tf.tensor1d(extractWeights(channelsOut))
+
+    paramMappings.push(
+      { paramPath: `${mappedPrefix}/weights` },
+      { paramPath: `${mappedPrefix}/bias` }
+    )
+
+    return {
+      weights,
+      bias
+    }
+  }
+
+  const extractConvParams = extractConvParamsFactory(extractWeights, paramMappings)
+
+  function extractDenseBlock3Params(channelsIn: number, channelsOut: number, mappedPrefix: string, isFirstLayer: boolean = false): DenseBlock3Params {
+
+    const conv0 = isFirstLayer
+      ? extractConvParams(channelsIn, channelsOut, 3, `${mappedPrefix}/conv0`)
+      : extractSeparableConvParams(channelsIn, channelsOut, `${mappedPrefix}/conv0`)
+    const conv1 = extractSeparableConvParams(channelsOut, channelsOut, `${mappedPrefix}/conv1`)
+    const conv2 = extractSeparableConvParams(channelsOut, channelsOut, `${mappedPrefix}/conv2`)
+
+    return { conv0, conv1, conv2 }
+  }
+
+  function extractDenseBlock4Params(channelsIn: number, channelsOut: number, mappedPrefix: string, isFirstLayer: boolean = false): DenseBlock4Params {
+
+    const { conv0, conv1, conv2 } = extractDenseBlock3Params(channelsIn, channelsOut, mappedPrefix, isFirstLayer)
+    const conv3 = extractSeparableConvParams(channelsOut, channelsOut, `${mappedPrefix}/conv3`)
+
+    return { conv0, conv1, conv2, conv3 }
+  }
+
+  return {
+    extractDenseBlock3Params,
+    extractDenseBlock4Params,
+    extractFCParams
+  }
+
+}
\ No newline at end of file

src/faceLandmarkNet/index.ts

 import { FaceLandmark68Net } from './FaceLandmark68Net';
 
 export * from './FaceLandmark68Net';
+export * from './FaceLandmark68TinyNet';
 
 export class FaceLandmarkNet extends FaceLandmark68Net {}
 

src/faceLandmarkNet/loadParamsFactory.ts

+import * as tf from '@tensorflow/tfjs-core';
+import { extractWeightEntryFactory, ParamMapping } from 'tfjs-image-recognition-base';
+import { ConvParams, FCParams } from 'tfjs-tiny-yolov2';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
+
+import { DenseBlock3Params, DenseBlock4Params } from './types';
+
+export function loadParamsFactory(weightMap: any, paramMappings: ParamMapping[]) {
+
+  const extractWeightEntry = extractWeightEntryFactory(weightMap, paramMappings)
+
+  function extractConvParams(prefix: string): ConvParams {
+    const filters = extractWeightEntry<tf.Tensor4D>(`${prefix}/filters`, 4)
+    const bias = extractWeightEntry<tf.Tensor1D>(`${prefix}/bias`, 1)
+
+    return { filters, bias }
+  }
+
+  function extractSeparableConvParams(prefix: string): SeparableConvParams {
+    const depthwise_filter = extractWeightEntry<tf.Tensor4D>(`${prefix}/depthwise_filter`, 4)
+    const pointwise_filter = extractWeightEntry<tf.Tensor4D>(`${prefix}/pointwise_filter`, 4)
+    const bias = extractWeightEntry<tf.Tensor1D>(`${prefix}/bias`, 1)
+
+    return new SeparableConvParams(
+      depthwise_filter,
+      pointwise_filter,
+      bias
+    )
+  }
+
+  function extractDenseBlock3Params(prefix: string, isFirstLayer: boolean = false): DenseBlock3Params {
+    const conv0 = isFirstLayer
+      ? extractConvParams(`${prefix}/conv0`)
+      : extractSeparableConvParams(`${prefix}/conv0`)
+    const conv1 = extractSeparableConvParams(`${prefix}/conv1`)
+    const conv2 = extractSeparableConvParams(`${prefix}/conv2`)
+
+    return { conv0, conv1, conv2 }
+  }
+
+  function extractDenseBlock4Params(prefix: string, isFirstLayer: boolean = false): DenseBlock4Params {
+    const conv0 = isFirstLayer
+      ? extractConvParams(`${prefix}/conv0`)
+      : extractSeparableConvParams(`${prefix}/conv0`)
+    const conv1 = extractSeparableConvParams(`${prefix}/conv1`)
+    const conv2 = extractSeparableConvParams(`${prefix}/conv2`)
+    const conv3 = extractSeparableConvParams(`${prefix}/conv3`)
+
+    return { conv0, conv1, conv2, conv3 }
+  }
+
+  function extractFcParams(prefix: string): FCParams {
+    const weights = extractWeightEntry<tf.Tensor2D>(`${prefix}/weights`, 2)
+    const bias = extractWeightEntry<tf.Tensor1D>(`${prefix}/bias`, 1)
+
+    return { weights, bias }
+  }
+
+  return {
+    extractDenseBlock3Params,
+    extractDenseBlock4Params,
+    extractFcParams
+  }
+}
\ No newline at end of file

src/faceLandmarkNet/loadQuantizedParams.ts

-import * as tf from '@tensorflow/tfjs-core';
-import {
-  disposeUnusedWeightTensors,
-  extractWeightEntryFactory,
-  loadWeightMap,
-  ParamMapping,
-} from 'tfjs-image-recognition-base';
-import { ConvParams, FCParams } from 'tfjs-tiny-yolov2';
+import { disposeUnusedWeightTensors, loadWeightMap, ParamMapping } from 'tfjs-image-recognition-base';
 
+import { loadParamsFactory } from './loadParamsFactory';
 import { NetParams } from './types';
 
 const DEFAULT_MODEL_NAME = 'face_landmark_68_model'
 
-function extractorsFactory(weightMap: any, paramMappings: ParamMapping[]) {
-
-  const extractWeightEntry = extractWeightEntryFactory(weightMap, paramMappings)
-
-  function extractConvParams(prefix: string, mappedPrefix: string): ConvParams {
-    const filters = extractWeightEntry<tf.Tensor4D>(`${prefix}/kernel`, 4, `${mappedPrefix}/filters`)
-    const bias = extractWeightEntry<tf.Tensor1D>(`${prefix}/bias`, 1, `${mappedPrefix}/bias`)
-
-    return { filters, bias }
-  }
-
-  function extractFcParams(prefix: string, mappedPrefix: string): FCParams {
-    const weights = extractWeightEntry<tf.Tensor2D>(`${prefix}/kernel`, 2, `${mappedPrefix}/weights`)
-    const bias = extractWeightEntry<tf.Tensor1D>(`${prefix}/bias`, 1, `${mappedPrefix}/bias`)
-
-    return { weights, bias }
-  }
-
-  return {
-    extractConvParams,
-    extractFcParams
-  }
-}
-
 export async function loadQuantizedParams(
   uri: string | undefined
 ): Promise<{ params: NetParams, paramMappings: ParamMapping[] }> {
@@ -43,21 +13,16 @@ export async function loadQuantizedParams(
   const paramMappings: ParamMapping[] = []
 
   const {
-    extractConvParams,
+    extractDenseBlock4Params,
     extractFcParams
-  } = extractorsFactory(weightMap, paramMappings)
+  } = loadParamsFactory(weightMap, paramMappings)
 
   const params = {
-    conv0: extractConvParams('conv2d_0', 'conv0'),
-    conv1: extractConvParams('conv2d_1', 'conv1'),
-    conv2: extractConvParams('conv2d_2', 'conv2'),
-    conv3: extractConvParams('conv2d_3', 'conv3'),
-    conv4: extractConvParams('conv2d_4', 'conv4'),
-    conv5: extractConvParams('conv2d_5', 'conv5'),
-    conv6: extractConvParams('conv2d_6', 'conv6'),
-    conv7: extractConvParams('conv2d_7', 'conv7'),
-    fc0: extractFcParams('dense', 'fc0'),
-    fc1: extractFcParams('logits', 'fc1')
+    dense0: extractDenseBlock4Params('dense0', true),
+    dense1: extractDenseBlock4Params('dense1'),
+    dense2: extractDenseBlock4Params('dense2'),
+    dense3: extractDenseBlock4Params('dense3'),
+    fc: extractFcParams('fc')
   }
 
   disposeUnusedWeightTensors(weightMap, paramMappings)

src/faceLandmarkNet/loadQuantizedParamsTiny.ts

+import { disposeUnusedWeightTensors, loadWeightMap, ParamMapping } from 'tfjs-image-recognition-base';
+
+import { loadParamsFactory } from './loadParamsFactory';
+import { TinyNetParams } from './types';
+
+const DEFAULT_MODEL_NAME = 'face_landmark_68_tiny_model'
+
+export async function loadQuantizedParamsTiny(
+  uri: string | undefined
+): Promise<{ params: TinyNetParams, paramMappings: ParamMapping[] }> {
+
+  const weightMap = await loadWeightMap(uri, DEFAULT_MODEL_NAME)
+  const paramMappings: ParamMapping[] = []
+
+  const {
+    extractDenseBlock3Params,
+    extractFcParams
+  } = loadParamsFactory(weightMap, paramMappings)
+
+  const params = {
+    dense0: extractDenseBlock3Params('dense0', true),
+    dense1: extractDenseBlock3Params('dense1'),
+    dense2: extractDenseBlock3Params('dense2'),
+    fc: extractFcParams('fc')
+  }
+
+  disposeUnusedWeightTensors(weightMap, paramMappings)
+
+  return { params, paramMappings }
+}
\ No newline at end of file

src/faceLandmarkNet/types.ts

+import * as tf from '@tensorflow/tfjs-core';
 import { ConvParams, FCParams } from 'tfjs-tiny-yolov2';
+import { SeparableConvParams } from 'tfjs-tiny-yolov2/build/tinyYolov2/types';
+
+export type ConvWithBatchNormParams = BatchNormParams & {
+  filter: tf.Tensor4D
+}
+
+export type BatchNormParams = {
+  mean: tf.Tensor1D
+  variance: tf.Tensor1D
+  scale: tf.Tensor1D
+  offset: tf.Tensor1D
+}
+
+export type SeparableConvWithBatchNormParams = {
+  depthwise: ConvWithBatchNormParams
+  pointwise: ConvWithBatchNormParams
+}
+
+export declare type FCWithBatchNormParams = BatchNormParams & {
+    weights: tf.Tensor2D
+}
+
+export type DenseBlock3Params = {
+  conv0: SeparableConvParams | ConvParams
+  conv1: SeparableConvParams
+  conv2: SeparableConvParams
+}
+
+export type DenseBlock4Params = DenseBlock3Params & {
+  conv3: SeparableConvParams
+}
+
+export type TinyNetParams = {
+  dense0: DenseBlock3Params
+  dense1: DenseBlock3Params
+  dense2: DenseBlock3Params
+  fc: FCParams
+}
 
 export type NetParams = {
-  conv0: ConvParams
-  conv1: ConvParams
-  conv2: ConvParams
-  conv3: ConvParams
-  conv4: ConvParams
-  conv5: ConvParams
-  conv6: ConvParams
-  conv7: ConvParams
-  fc0: FCParams
-  fc1: FCParams
-}
\ No newline at end of file
+  dense0: DenseBlock4Params
+  dense1: DenseBlock4Params
+  dense2: DenseBlock4Params
+  dense3: DenseBlock4Params
+  fc: FCParams
+}
+

src/globalApi.ts

@@ -8,6 +8,7 @@ import { FaceLandmarks68 } from './classes/FaceLandmarks68';
 import { FullFaceDescription } from './classes/FullFaceDescription';
 import { FaceDetectionNet } from './faceDetectionNet/FaceDetectionNet';
 import { FaceLandmark68Net } from './faceLandmarkNet/FaceLandmark68Net';
+import { FaceLandmark68TinyNet } from './faceLandmarkNet/FaceLandmark68TinyNet';
 import { FaceRecognitionNet } from './faceRecognitionNet/FaceRecognitionNet';
 import { Mtcnn } from './mtcnn/Mtcnn';
 import { MtcnnForwardParams, MtcnnResult } from './mtcnn/types';
@@ -22,6 +23,7 @@ export const recognitionNet = new FaceRecognitionNet()
 export const nets = {
   ssdMobilenetv1: detectionNet,
   faceLandmark68Net: landmarkNet,
+  faceLandmark68TinyNet: new FaceLandmark68TinyNet(),
   faceRecognitionNet: recognitionNet,
   mtcnn: new Mtcnn(),
   tinyYolov2: new TinyYolov2()
@@ -35,6 +37,10 @@ export function loadFaceLandmarkModel(url: string) {
   return nets.faceLandmark68Net.load(url)
 }
 
+export function loadFaceLandmarkTinyModel(url: string) {
+  return nets.faceLandmark68TinyNet.load(url)
+}
+
 export function loadFaceRecognitionModel(url: string) {
   return nets.faceRecognitionNet.load(url)
 }
@@ -52,6 +58,7 @@ export function loadFaceDetectionModel(url: string) {
 }
 
 export function loadModels(url: string) {
+  console.warn('loadModels will be deprecated in future')
   return Promise.all([
     loadSsdMobilenetv1Model(url),
     loadFaceLandmarkModel(url),
@@ -76,6 +83,11 @@ export function detectLandmarks(
 ): Promise<FaceLandmarks68 | FaceLandmarks68[]> {
   return nets.faceLandmark68Net.detectLandmarks(input)
 }
+export function detectLandmarksTiny(
+  input: TNetInput
+): Promise<FaceLandmarks68 | FaceLandmarks68[]> {
+  return nets.faceLandmark68TinyNet.detectLandmarks(input)
+}
 
 export function computeFaceDescriptor(
   input: TNetInput

test/data/faceLandmarkPositions1.json

-[{"x": 15.145603939890862, "y": 42.00849384069443}, {"x": 16.228551417589188, "y": 58.02471041679382}, {"x": 18.641504645347595, "y": 73.78255426883698}, {"x": 21.99183627963066, "y": 89.28244113922119}, {"x": 27.241109311580658, "y": 103.89149487018585}, {"x": 35.77959090471268, "y": 117.20520257949829}, {"x": 47.162988781929016, "y": 128.52083444595337}, {"x": 60.60639023780823, "y": 137.3870998620987}, {"x": 76.22466087341309, "y": 139.8476779460907}, {"x": 91.47888422012329, "y": 136.8298441171646}, {"x": 104.65206205844879, "y": 127.64202654361725}, {"x": 115.92219471931458, "y": 116.16075038909912}, {"x": 124.51052069664001, "y": 102.82219648361206}, {"x": 129.76784706115723, "y": 87.81199157238007}, {"x": 132.6519101858139, "y": 72.15077430009842}, {"x": 134.51203107833862, "y": 56.30408674478531}, {"x": 135.33487915992737, "y": 40.10292738676071}, {"x": 24.932710826396942, "y": 32.96273946762085}, {"x": 32.57601857185364, "y": 28.00302878022194}, {"x": 42.381374537944794, "y": 27.10871994495392}, {"x": 52.198585867881775, "y": 29.124750941991806}, {"x": 61.5922674536705, "y": 32.90563523769379}, {"x": 85.18133461475372, "y": 31.272751092910767}, {"x": 94.89730596542358, "y": 26.85772553086281}, {"x": 105.1171749830246, "y": 24.783611297607422}, {"x": 115.2367115020752, "y": 25.725752860307693}, {"x": 123.13878536224365, "y": 30.079219490289688}, {"x": 73.67254793643951, "y": 43.12345236539841}, {"x": 73.7981915473938, "y": 55.55519610643387}, {"x": 73.8839328289032, "y": 67.69467741250992}, {"x": 73.96439015865326, "y": 79.91722226142883}, {"x": 62.394945323467255, "y": 85.0008487701416}, {"x": 68.01678389310837, "y": 87.24332749843597}, {"x": 74.27898198366165, "y": 88.8382226228714}, {"x": 80.42970299720764, "y": 86.84954345226288}, {"x": 85.87614297866821, "y": 84.31319296360016}, {"x": 35.7684463262558, "y": 44.57894414663315}, {"x": 42.30532944202423, "y": 41.117119789123535}, {"x": 50.501713156700134, "y": 41.50397926568985}, {"x": 56.56731426715851, "y": 45.83812057971954}, {"x": 50.084468722343445, "y": 48.037388920784}, {"x": 41.91972613334656, "y": 47.92611300945282}, {"x": 91.28680229187012, "y": 44.677162170410156}, {"x": 97.76994287967682, "y": 40.084308385849}, {"x": 105.72623312473297, "y": 39.8311972618103}, {"x": 111.95003092288971, "y": 42.97384321689606}, {"x": 106.4166247844696, "y": 46.30742222070694}, {"x": 98.37331473827362, "y": 46.70642763376236}, {"x": 51.94934159517288, "y": 102.49188244342804}, {"x": 60.71081757545471, "y": 101.18674635887146}, {"x": 68.67328137159348, "y": 100.15718936920166}, {"x": 74.38371777534485, "y": 101.07522904872894}, {"x": 80.5251270532608, "y": 99.71560835838318}, {"x": 88.56176733970642, "y": 100.32560527324677}, {"x": 96.980881690979, "y": 100.81177353858948}, {"x": 89.01146650314331, "y": 108.55080485343933}, {"x": 81.3762366771698, "y": 112.2200220823288}, {"x": 74.73004907369614, "y": 113.34742605686188}, {"x": 68.42194050550461, "y": 112.92429864406586}, {"x": 60.53387671709061, "y": 110.02114713191986}, {"x": 55.876149237155914, "y": 103.3714234828949}, {"x": 68.54872852563858, "y": 104.03825640678406}, {"x": 74.44304376840591, "y": 104.46503162384033}, {"x": 80.63703775405884, "y": 103.56011688709259}, {"x": 93.1486040353775, "y": 102.0165503025055}, {"x": 80.84633946418762, "y": 104.64600920677185}, {"x": 74.63692724704742, "y": 105.62551617622375}, {"x": 68.74832957983017, "y": 105.22661805152893}]
\ No newline at end of file
+[{"x":12.409065663814545,"y":35.459357500076294},{"x":13.863831013441086,"y":53.03075462579727},{"x":17.257214337587357,"y":69.21994239091873},{"x":20.290792733430862,"y":83.73786807060242},{"x":25.19843727350235,"y":100.01502335071564},{"x":33.363962173461914,"y":114.15562033653259},{"x":42.67177730798721,"y":124.92083609104156},{"x":54.774945974349976,"y":134.97415781021118},{"x":73.93101006746292,"y":140.77884256839752},{"x":93.50163638591766,"y":135.3241503238678},{"x":106.89235031604767,"y":125.59508979320526},{"x":116.92832708358765,"y":115.29875993728638},{"x":125.80114603042603,"y":101.27357840538025},{"x":130.48838675022125,"y":84.83449816703796},{"x":133.11561048030853,"y":69.95338797569275},{"x":135.84416806697845,"y":53.477123379707336},{"x":137.16664016246796,"y":35.98061129450798},{"x":22.041092813014984,"y":26.081138849258423},{"x":29.21232059597969,"y":21.911372244358063},{"x":38.767728209495544,"y":21.422917395830154},{"x":47.85865992307663,"y":23.588016629219055},{"x":55.94828277826309,"y":26.723874360322952},{"x":88.98929357528687,"y":26.476483047008514},{"x":97.3086029291153,"y":23.411455750465393},{"x":106.5034031867981,"y":21.319837868213654},{"x":116.54528975486755,"y":21.732238680124283},{"x":124.39115345478058,"y":26.091013848781586},{"x":72.58472889661789,"y":44.89734023809433},{"x":72.1238225698471,"y":57.7533558011055},{"x":71.73694968223572,"y":70.3047975897789},{"x":71.81699573993683,"y":79.94250655174255},{"x":62.10947334766388,"y":83.842733502388},{"x":66.3440153002739,"y":85.69430708885193},{"x":72.36149311065674,"y":87.41167187690735},{"x":78.35066020488739,"y":85.74668169021606},{"x":83.0715000629425,"y":83.87293517589569},{"x":33.77652391791344,"y":40.25426656007767},{"x":39.713045954704285,"y":37.594060599803925},{"x":47.61984497308731,"y":37.72639185190201},{"x":55.496129393577576,"y":42.38396733999252},{"x":48.19572865962982,"y":44.55205053091049},{"x":39.7584468126297,"y":44.11860108375549},{"x":90.26297628879547,"y":42.310331761837006},{"x":97.85987734794617,"y":37.882326543331146},{"x":105.9216856956482,"y":37.697312235832214},{"x":112.38531768321991,"y":40.45048803091049},{"x":105.98326027393341,"y":44.305725395679474},{"x":97.26411402225494,"y":44.69909369945526},{"x":50.48552602529526,"y":100.74669420719147},{"x":58.16384553909302,"y":99.99362826347351},{"x":67.69992560148239,"y":98.86696636676788},{"x":72.78527766466141,"y":100.00039637088776},{"x":77.79269814491272,"y":98.80055487155914},{"x":88.17337453365326,"y":99.85068440437317},{"x":97.4126547574997,"y":100.27143359184265},{"x":87.91132271289825,"y":107.21607506275177},{"x":80.38677871227264,"y":111.03568375110626},{"x":73.13570827245712,"y":111.78229451179504},{"x":65.85598737001419,"y":111.23852133750916},{"x":58.75929147005081,"y":107.56462812423706},{"x":52.198053896427155,"y":100.71471333503723},{"x":66.12485200166702,"y":103.38914394378662},{"x":72.73153960704803,"y":103.6908209323883},{"x":79.6861320734024,"y":103.28285694122314},{"x":95.68716287612915,"y":100.31683444976807},{"x":79.67764735221863,"y":103.91019880771637},{"x":72.93425649404526,"y":104.61705923080444},{"x":66.34287983179092,"y":103.94268035888672}]
\ No newline at end of file

test/data/faceLandmarkPositions1Tiny.json

+[{"x":12.591811269521713,"y":37.18056008219719},{"x":14.489217102527618,"y":54.375430941581726},{"x":18.031097576022148,"y":70.10978311300278},{"x":21.278046816587448,"y":83.95282924175262},{"x":26.40417516231537,"y":99.72459375858307},{"x":34.79916527867317,"y":113.51814866065979},{"x":44.21398043632507,"y":124.09239113330841},{"x":56.03765398263931,"y":134.3254119157791},{"x":75.41389167308807,"y":140.46012461185455},{"x":94.84915137290955,"y":134.49379205703735},{"x":107.92432129383087,"y":123.9702969789505},{"x":118.00838112831116,"y":113.45937252044678},{"x":126.69345438480377,"y":99.56673681735992},{"x":131.41701221466064,"y":83.40703547000885},{"x":133.99315774440765,"y":68.92484575510025},{"x":136.772882938385,"y":52.720874547958374},{"x":137.7272218465805,"y":35.40073335170746},{"x":21.938379853963852,"y":28.243593871593475},{"x":29.08402383327484,"y":24.014483392238617},{"x":38.46672624349594,"y":23.24410155415535},{"x":47.45165705680847,"y":24.920182675123215},{"x":55.57771772146225,"y":27.864713966846466},{"x":89.10931348800659,"y":27.321801334619522},{"x":97.32098579406738,"y":24.548690021038055},{"x":106.52551352977753,"y":22.409027069807053},{"x":116.49984419345856,"y":22.863386571407318},{"x":124.49341714382172,"y":26.976196467876434},{"x":72.84769266843796,"y":46.6731995344162},{"x":72.5912556052208,"y":59.78170037269592},{"x":72.49848246574402,"y":72.47795909643173},{"x":72.59156405925751,"y":82.31859505176544},{"x":63.02006542682648,"y":85.39192378520966},{"x":67.24515706300735,"y":87.40304410457611},{"x":73.31620305776596,"y":89.13427591323853},{"x":79.3295681476593,"y":87.3850017786026},{"x":83.98161828517914,"y":85.33579409122467},{"x":34.017493575811386,"y":42.064668238162994},{"x":39.69891518354416,"y":39.04064744710922},{"x":47.665129601955414,"y":39.02306109666824},{"x":55.84650635719299,"y":43.793052434921265},{"x":48.43880832195282,"y":46.189695596694946},{"x":39.91182893514633,"y":45.974622666835785},{"x":90.23652970790863,"y":43.494392931461334},{"x":97.89576530456543,"y":38.67722600698471},{"x":105.97457885742188,"y":38.45117390155792},{"x":112.57963478565216,"y":41.22479259967804},{"x":106.20929896831512,"y":45.47332227230072},{"x":97.40135371685028,"y":45.80201804637909},{"x":51.17177367210388,"y":100.92100203037262},{"x":58.8924765586853,"y":100.64716637134552},{"x":68.65793466567993,"y":99.5934247970581},{"x":73.81858080625534,"y":100.83856880664825},{"x":78.77743542194366,"y":99.58769381046295},{"x":89.21037912368774,"y":100.49387812614441},{"x":98.3612447977066,"y":100.58583319187164},{"x":88.92414271831512,"y":107.61089622974396},{"x":81.55436217784882,"y":111.79174482822418},{"x":74.08711463212967,"y":112.78842687606812},{"x":66.74499660730362,"y":112.13003396987915},{"x":59.4507560133934,"y":108.14197361469269},{"x":52.86548137664795,"y":100.84743797779083},{"x":67.18341708183289,"y":104.20102179050446},{"x":73.75064492225647,"y":104.48753535747528},{"x":80.83339333534241,"y":104.14700210094452},{"x":96.58516645431519,"y":100.67276358604431},{"x":80.59312999248505,"y":104.5551985502243},{"x":73.77402484416962,"y":105.33222556114197},{"x":67.1465814113617,"y":104.65545952320099}]
\ No newline at end of file

test/data/faceLandmarkPositions2.json

-[{"x": 9.995004907250404, "y": 53.55449616909027}, {"x": 12.50796876847744, "y": 71.41348421573639}, {"x": 16.677917540073395, "y": 88.59677910804749}, {"x": 22.6475290954113, "y": 104.6014130115509}, {"x": 30.59161528944969, "y": 119.35952603816986}, {"x": 41.422560811042786, "y": 132.23226964473724}, {"x": 54.74700182676315, "y": 142.4335777759552}, {"x": 70.32481580972672, "y": 149.33189749717712}, {"x": 87.31497824192047, "y": 150.50972700119019}, {"x": 103.98584604263306, "y": 145.98273038864136}, {"x": 117.90181696414948, "y": 135.19554734230042}, {"x": 128.67935299873352, "y": 121.79077863693237}, {"x": 136.7296814918518, "y": 105.85636496543884}, {"x": 140.29521346092224, "y": 88.25878500938416}, {"x": 140.9232795238495, "y": 70.16736567020416}, {"x": 140.2374029159546, "y": 52.73242145776749}, {"x": 137.97148168087006, "y": 34.537942707538605}, {"x": 14.37721811234951, "y": 33.1049881875515}, {"x": 22.6781465113163, "y": 24.685607850551605}, {"x": 34.36600640416145, "y": 21.1758591234684}, {"x": 46.24761343002319, "y": 22.49436378479004}, {"x": 57.12086856365204, "y": 26.742971688508987}, {"x": 81.21025264263153, "y": 23.014162480831146}, {"x": 92.2086775302887, "y": 15.48520065844059}, {"x": 104.77548837661743, "y": 11.306393891572952}, {"x": 117.67798662185669, "y": 11.740228906273842}, {"x": 127.28274464607239, "y": 18.115675449371338}, {"x": 69.62742805480957, "y": 41.51403307914734}, {"x": 70.82946002483368, "y": 55.146731436252594}, {"x": 71.84555232524872, "y": 68.59723627567291}, {"x": 73.0046421289444, "y": 81.93029165267944}, {"x": 60.417647659778595, "y": 88.01697492599487}, {"x": 67.98770427703857, "y": 90.65443575382233}, {"x": 76.07284784317017, "y": 91.86699986457825}, {"x": 84.35145914554596, "y": 88.2117748260498}, {"x": 90.86072444915771, "y": 83.67109894752502}, {"x": 28.828849643468857, "y": 47.794362902641296}, {"x": 36.311765760183334, "y": 43.33548992872238}, {"x": 44.95347887277603, "y": 43.20283681154251}, {"x": 52.85406410694122, "y": 48.07424694299698}, {"x": 44.7566494345665, "y": 49.9691516160965}, {"x": 35.997654497623444, "y": 50.32083839178085}, {"x": 89.51361179351807, "y": 42.501528561115265}, {"x": 97.55686819553375, "y": 35.38782298564911}, {"x": 106.73499405384064, "y": 33.59129726886749}, {"x": 114.8474782705307, "y": 36.34611591696739}, {"x": 108.40394496917725, "y": 40.97002297639847}, {"x": 98.98389279842377, "y": 42.47862249612808}, {"x": 53.17014008760452, "y": 109.99322533607483}, {"x": 62.47727572917938, "y": 105.5664449930191}, {"x": 72.82306104898453, "y": 102.35638618469238}, {"x": 80.85319697856903, "y": 103.16510796546936}, {"x": 89.42103087902069, "y": 100.08856952190399}, {"x": 99.8135894536972, "y": 100.0435084104538}, {"x": 109.78849232196808, "y": 101.60946249961853}, {"x": 101.90783143043518, "y": 115.25328755378723}, {"x": 92.65078604221344, "y": 122.49988317489624}, {"x": 83.28675627708435, "y": 125.00075697898865}, {"x": 74.31002408266068, "y": 125.16917288303375}, {"x": 63.37641924619675, "y": 121.38420939445496}, {"x": 57.85811394453049, "y": 110.28821468353271}, {"x": 73.27612638473511, "y": 108.2253098487854}, {"x": 81.34059011936188, "y": 108.251291513443}, {"x": 89.9710088968277, "y": 105.99507093429565}, {"x": 104.85810041427612, "y": 103.1228095293045}, {"x": 90.87785482406616, "y": 112.19902038574219}, {"x": 82.05846548080444, "y": 114.52528238296509}, {"x": 73.8232746720314, "y": 114.64338898658752}]
\ No newline at end of file
+[{"x":9.57956425845623,"y":52.39383280277252},{"x":12.517975643277168,"y":69.67290490865707},{"x":17.382895946502686,"y":85.59838235378265},{"x":22.254569828510284,"y":100.24277865886688},{"x":29.952556639909744,"y":115.30093252658844},{"x":40.607236325740814,"y":127.02118456363678},{"x":52.21455842256546,"y":135.75681746006012},{"x":66.94714576005936,"y":144.7398841381073},{"x":87.62003481388092,"y":148.519966006279},{"x":107.54483342170715,"y":139.5822286605835},{"x":120.26938498020172,"y":127.25696861743927},{"x":129.46229875087738,"y":116.31288528442383},{"x":136.60779297351837,"y":101.43148005008698},{"x":139.15124237537384,"y":84.36994850635529},{"x":139.66587781906128,"y":68.84740144014359},{"x":139.49655890464783,"y":52.11438238620758},{"x":138.19187879562378,"y":34.275998175144196},{"x":17.73911565542221,"y":34.247177839279175},{"x":23.905900865793228,"y":27.4917371571064},{"x":33.15091207623482,"y":23.992185294628143},{"x":42.38297492265701,"y":24.080870300531387},{"x":50.80330967903137,"y":25.668590515851974},{"x":82.72879421710968,"y":22.588082402944565},{"x":90.8599466085434,"y":18.972241133451462},{"x":100.4738599061966,"y":16.92224331200123},{"x":111.04867458343506,"y":17.707618698477745},{"x":120.33791542053223,"y":22.05384224653244},{"x":69.454425573349,"y":45.07347196340561},{"x":70.71190774440765,"y":59.0581476688385},{"x":71.70603275299072,"y":72.0032662153244},{"x":73.040771484375,"y":81.89877569675446},{"x":63.307446241378784,"y":87.61201500892639},{"x":68.28959137201309,"y":88.8374537229538},{"x":75.32510161399841,"y":89.66150879859924},{"x":81.8707287311554,"y":87.05066442489624},{"x":87.0687872171402,"y":84.28579866886139},{"x":31.470197439193726,"y":46.391429007053375},{"x":36.97886019945145,"y":42.39812046289444},{"x":45.00039964914322,"y":41.7159765958786},{"x":53.23628783226013,"y":45.5289289355278},{"x":46.301740407943726,"y":48.1160894036293},{"x":37.982627749443054,"y":48.80691468715668},{"x":87.06339597702026,"y":41.115787625312805},{"x":93.96767914295197,"y":35.74279770255089},{"x":102.04813778400421,"y":34.39944013953209},{"x":109.27990972995758,"y":36.874495446681976},{"x":102.8844952583313,"y":40.83576053380966},{"x":94.25925314426422,"y":42.34584420919418},{"x":51.57048851251602,"y":105.63550293445587},{"x":58.90470743179321,"y":101.6682118177414},{"x":70.55880725383759,"y":98.85237514972687},{"x":77.36653983592987,"y":99.13035929203033},{"x":83.67621302604675,"y":97.07690477371216},{"x":97.94619083404541,"y":96.29683792591095},{"x":109.81523394584656,"y":97.79501259326935},{"x":101.1863261461258,"y":112.63381540775299},{"x":92.58085191249847,"y":120.5494076013565},{"x":82.98978209495544,"y":123.54103624820709},{"x":73.21132868528366,"y":123.69250059127808},{"x":63.237668573856354,"y":118.47777664661407},{"x":53.231893479824066,"y":105.59755861759186},{"x":69.4526419043541,"y":103.36487889289856},{"x":78.06874215602875,"y":102.5748610496521},{"x":87.34851479530334,"y":101.16179287433624},{"x":108.10398459434509,"y":98.21300804615021},{"x":90.36072492599487,"y":113.51232826709747},{"x":81.56309723854065,"y":116.386878490448},{"x":72.68635332584381,"y":116.3683533668518}]
\ No newline at end of file

test/data/faceLandmarkPositions2Tiny.json

+[{"x":10.509434714913368,"y":52.461858093738556},{"x":13.533616438508034,"y":69.72497999668121},{"x":18.125666677951813,"y":85.49389243125916},{"x":22.68524095416069,"y":99.90644752979279},{"x":29.770859330892563,"y":115.02401232719421},{"x":40.009237825870514,"y":127.17496454715729},{"x":51.13186240196228,"y":134.8838210105896},{"x":64.93598967790604,"y":142.67239272594452},{"x":86.1398845911026,"y":145.5208271741867},{"x":106.27160668373108,"y":137.6600593328476},{"x":119.85296308994293,"y":126.55811011791229},{"x":129.32697236537933,"y":115.91286063194275},{"x":136.9260460138321,"y":101.54499113559723},{"x":140.04463255405426,"y":85.47771871089935},{"x":140.39686918258667,"y":70.350381731987},{"x":140.93535840511322,"y":53.595203161239624},{"x":139.76540565490723,"y":35.78364551067352},{"x":18.718457221984863,"y":33.92140567302704},{"x":24.8894065618515,"y":27.165759354829788},{"x":34.25554856657982,"y":23.889242112636566},{"x":43.19920241832733,"y":23.89446794986725},{"x":51.55668407678604,"y":25.493214279413223},{"x":83.17877948284149,"y":22.116592526435852},{"x":91.0528689622879,"y":19.1002294421196},{"x":100.33524334430695,"y":17.00422614812851},{"x":110.7668548822403,"y":18.07868331670761},{"x":119.95638906955719,"y":22.73145094513893},{"x":69.94039714336395,"y":44.93928104639053},{"x":70.8938866853714,"y":58.68610739707947},{"x":71.92910313606262,"y":71.54497504234314},{"x":73.07412475347519,"y":81.73994421958923},{"x":63.52762430906296,"y":87.66357600688934},{"x":68.80622059106827,"y":89.00657594203949},{"x":75.79838633537292,"y":89.81735408306122},{"x":82.58680701255798,"y":87.12102770805359},{"x":87.82331049442291,"y":84.34564769268036},{"x":32.21326768398285,"y":45.98721116781235},{"x":37.76122033596039,"y":42.323389649391174},{"x":45.626114308834076,"y":41.58268868923187},{"x":53.76068651676178,"y":45.14274448156357},{"x":46.830734610557556,"y":47.79057204723358},{"x":38.68361413478851,"y":48.451441526412964},{"x":87.2931182384491,"y":41.23439937829971},{"x":93.99429559707642,"y":36.358749121427536},{"x":101.83376669883728,"y":35.318271070718765},{"x":109.02988314628601,"y":37.269607186317444},{"x":102.72921323776245,"y":41.246773302555084},{"x":94.33983564376831,"y":42.437638342380524},{"x":51.489101350307465,"y":106.93963766098022},{"x":59.81220602989197,"y":103.0700147151947},{"x":71.77513092756271,"y":100.07118880748749},{"x":78.22524905204773,"y":100.49879550933838},{"x":84.36182141304016,"y":98.41773211956024},{"x":98.11981916427612,"y":97.42594063282013},{"x":109.53287780284882,"y":98.12568426132202},{"x":99.13961291313171,"y":110.45710444450378},{"x":90.27182757854462,"y":117.69201457500458},{"x":80.78381717205048,"y":120.45529782772064},{"x":71.48520648479462,"y":120.73915600776672},{"x":62.03146129846573,"y":116.77920520305634},{"x":53.2006099820137,"y":106.71007633209229},{"x":70.40069103240967,"y":104.98761534690857},{"x":78.83743643760681,"y":104.08617854118347},{"x":88.0838692188263,"y":102.68035233020782},{"x":107.74409472942352,"y":98.4467625617981},{"x":88.4322077035904,"y":110.55790185928345},{"x":79.5111358165741,"y":113.10129761695862},{"x":71.03536427021027,"y":113.31479251384735}]
\ No newline at end of file

test/data/faceLandmarkPositionsRect.json

-[{"x": 1.41040606983006, "y": 39.10264679789543}, {"x": 6.364097021520138, "y": 59.51536446809769}, {"x": 9.736957371234894, "y": 77.80020213127136}, {"x": 13.817122921347618, "y": 96.05930614471436}, {"x": 20.00808236002922, "y": 110.73724734783173}, {"x": 26.37610113620758, "y": 121.9759613275528}, {"x": 33.717534959316254, "y": 132.17582309246063}, {"x": 41.72355371713638, "y": 140.95428133010864}, {"x": 50.880420446395874, "y": 146.48704254627228}, {"x": 61.8572758436203, "y": 146.95429050922394}, {"x": 72.31115305423737, "y": 142.27578461170197}, {"x": 81.12127649784088, "y": 134.44229400157928}, {"x": 90.62704622745514, "y": 123.21754229068756}, {"x": 99.09188437461853, "y": 110.66853940486908}, {"x": 106.08789837360382, "y": 92.88722932338715}, {"x": 110.42981934547424, "y": 72.83273243904114}, {"x": 110.97252583503723, "y": 51.15715354681015}, {"x": 19.203872233629227, "y": 37.71118628978729}, {"x": 30.421346753835678, "y": 37.16629707813263}, {"x": 39.97739166021347, "y": 40.400769114494324}, {"x": 49.04381215572357, "y": 45.61803185939789}, {"x": 56.723132252693176, "y": 53.29851043224335}, {"x": 56.084791123867035, "y": 46.76303869485855}, {"x": 65.71077406406403, "y": 46.11447751522064}, {"x": 75.0335134267807, "y": 46.96638810634613}, {"x": 83.45012521743774, "y": 51.2211879491806}, {"x": 90.43923103809357, "y": 59.06730401515961}, {"x": 57.634793639183044, "y": 60.69884181022644}, {"x": 56.081126153469086, "y": 72.10934686660767}, {"x": 54.33526223897934, "y": 84.50968235731125}, {"x": 52.046833634376526, "y": 99.18286514282227}, {"x": 43.99328297376633, "y": 99.4484453201294}, {"x": 49.026043593883514, "y": 104.39998996257782}, {"x": 55.01739031076431, "y": 107.4150230884552}, {"x": 59.81684720516205, "y": 105.90647208690643}, {"x": 64.0739357471466, "y": 101.74840021133423}, {"x": 27.087938010692596, "y": 55.09133046865463}, {"x": 34.27653384208679, "y": 55.21461433172226}, {"x": 42.54729217290878, "y": 57.39108908176422}, {"x": 48.454275488853455, "y": 61.82331371307373}, {"x": 41.1783966422081, "y": 61.66855639219284}, {"x": 34.177026987075806, "y": 60.50751310586929}, {"x": 59.51743584871292, "y": 62.6235288977623}, {"x": 65.59782898426056, "y": 61.08310657739639}, {"x": 73.38363683223724, "y": 63.624403953552246}, {"x": 78.78980994224548, "y": 67.99472141265869}, {"x": 72.09050583839417, "y": 68.4499539732933}, {"x": 64.50075209140778, "y": 66.9551106095314}, {"x": 34.73506963253021, "y": 112.76923489570618}, {"x": 42.499454855918884, "y": 113.63057744503021}, {"x": 49.24748706817627, "y": 114.29424142837524}, {"x": 55.50651115179062, "y": 115.7518025636673}, {"x": 62.20475721359253, "y": 116.19831931591034}, {"x": 66.55372452735901, "y": 117.66684019565582}, {"x": 71.82598805427551, "y": 118.23190891742706}, {"x": 64.1076112985611, "y": 127.00456368923187}, {"x": 58.63429069519043, "y": 129.7624831199646}, {"x": 52.65244817733765, "y": 129.39003217220306}, {"x": 47.463079154491425, "y": 127.39845669269562}, {"x": 42.28087282180786, "y": 123.10187613964081}, {"x": 38.320072412490845, "y": 114.88845479488373}, {"x": 49.383214592933655, "y": 119.21215355396271}, {"x": 55.4133198261261, "y": 120.68333745002747}, {"x": 60.75339984893799, "y": 121.00764191150665}, {"x": 67.49378943443298, "y": 118.56007528305054}, {"x": 59.50732082128525, "y": 124.34931361675262}, {"x": 54.06349813938141, "y": 124.39959263801575}, {"x": 48.627867102622986, "y": 122.32591986656189}]
\ No newline at end of file
+[{"x":9.103168368339539,"y":101.01195418834686},{"x":14.784232437610626,"y":109.96735978126526},{"x":20.72097545862198,"y":120.9920563697815},{"x":23.68257150053978,"y":127.4499876499176},{"x":33.172756016254425,"y":135.1237609386444},{"x":40.30949687957764,"y":136.57750260829926},{"x":46.08676642179489,"y":136.41320192813873},{"x":48.44881075620651,"y":143.47557425498962},{"x":58.82970488071442,"y":146.31987726688385},{"x":65.03185284137726,"y":143.93176698684692},{"x":70.69711172580719,"y":141.56521546840668},{"x":73.59640145301819,"y":140.66094636917114},{"x":85.81815421581268,"y":125.65490245819092},{"x":91.43581211566925,"y":117.32116794586182},{"x":93.22071814537048,"y":103.55141079425812},{"x":97.05579698085785,"y":97.57390463352203},{"x":89.83693969249725,"y":80.89369875192642},{"x":8.929109543561935,"y":95.41884541511536},{"x":15.941815882921219,"y":91.92907392978668},{"x":22.651149570941925,"y":87.75521492958069},{"x":25.080495923757553,"y":87.33296239376068},{"x":35.09866887331009,"y":87.67910146713257},{"x":55.91999834775925,"y":77.73741436004639},{"x":60.79969918727875,"y":76.88318556547165},{"x":72.479785323143,"y":77.81150668859482},{"x":72.09985733032227,"y":80.98973709344864},{"x":84.31956672668457,"y":76.90173119306564},{"x":46.01968437433243,"y":93.13016855716705},{"x":44.01630175113678,"y":100.97888398170471},{"x":40.80449330806732,"y":107.36412870883942},{"x":38.621174454689026,"y":114.48911905288696},{"x":33.12944173812866,"y":118.61227464675903},{"x":35.98146861791611,"y":119.76905286312103},{"x":43.14035201072693,"y":121.47148418426514},{"x":49.96515953540802,"y":119.40146112442017},{"x":54.14790153503418,"y":118.55489778518677},{"x":22.290891706943512,"y":101.89636206626892},{"x":26.193063497543335,"y":97.70843470096588},{"x":32.05891406536102,"y":95.37641751766205},{"x":36.877059400081635,"y":95.59315097332001},{"x":31.79427993297577,"y":98.80662655830383},{"x":26.30974444746971,"y":101.54476964473724},{"x":53.97727316617966,"y":88.98845142126083},{"x":59.00451451539993,"y":90.48788344860077},{"x":65.20665884017944,"y":89.76256692409515},{"x":72.8481148481369,"y":81.40269559621811},{"x":68.15013182163239,"y":88.9257538318634},{"x":59.68368661403656,"y":92.8312953710556},{"x":26.978514283895493,"y":124.41534793376923},{"x":37.64959371089935,"y":129.4168745279312},{"x":38.14149236679077,"y":127.94919180870056},{"x":44.06257927417755,"y":127.01305139064789},{"x":48.707089364528656,"y":122.96710205078125},{"x":57.109412133693695,"y":125.86821961402893},{"x":63.49719738960266,"y":116.9522500038147},{"x":51.657943189144135,"y":119.83492803573608},{"x":46.5363906621933,"y":126.18525648117065},{"x":43.10889798402786,"y":128.2001837491989},{"x":42.599557995796204,"y":130.2384203672409},{"x":31.888991057872772,"y":126.56950533390045},{"x":27.653676003217697,"y":126.88065385818481},{"x":44.846410274505615,"y":133.13909232616425},{"x":46.87078648805618,"y":129.8393074274063},{"x":52.942511677742004,"y":126.64793169498444},{"x":62.26701807975769,"y":119.98195624351501},{"x":47.63777244091034,"y":122.5474488735199},{"x":42.832436323165894,"y":123.82451903820038},{"x":39.61905354261398,"y":124.0300487279892}]
\ No newline at end of file

test/data/faceLandmarkPositionsRectTiny.json

+[{"x":28.1980222761631,"y":82.78948521614075},{"x":41.05900514125824,"y":94.1000794172287},{"x":45.85652047395706,"y":100.0026285648346},{"x":46.66536271572113,"y":103.72171652317047},{"x":46.17695450782776,"y":108.95469212532043},{"x":44.9439537525177,"y":113.12937867641449},{"x":41.77620494365692,"y":118.82230281829834},{"x":36.76300173997879,"y":120.5938982963562},{"x":40.35541081428528,"y":126.7483412027359},{"x":45.805581748485565,"y":129.04026460647583},{"x":52.93079322576523,"y":126.84588611125946},{"x":53.27251535654068,"y":126.36341333389282},{"x":55.39190810918808,"y":125.28188920021057},{"x":61.80034518241882,"y":123.59300637245178},{"x":70.61709320545197,"y":115.20878064632416},{"x":82.92120599746704,"y":111.2279531955719},{"x":81.62196671962738,"y":100.77173101902008},{"x":24.95584148168564,"y":76.60648649930954},{"x":32.25624072551727,"y":74.50906938314438},{"x":29.986725986003876,"y":73.17802804708481},{"x":37.48139047622681,"y":68.97298204898834},{"x":43.13521087169647,"y":72.66015625},{"x":66.04641699790955,"y":72.61856651306152},{"x":67.14672982692719,"y":77.07855653762817},{"x":75.16640496253967,"y":78.80701857805252},{"x":78.36494970321655,"y":81.77029263973236},{"x":83.96591889858246,"y":86.9630047082901},{"x":49.43720746040344,"y":86.65960651636124},{"x":45.592297196388245,"y":89.74523079395294},{"x":38.21692967414856,"y":95.09447729587555},{"x":36.639003574848175,"y":99.00447487831116},{"x":38.06128841638565,"y":99.63473975658417},{"x":38.76173406839371,"y":100.46630108356476},{"x":41.303012907505035,"y":101.67539536952972},{"x":44.19621616601944,"y":103.10262358188629},{"x":43.459695279598236,"y":107.30739903450012},{"x":35.52263468503952,"y":84.4736732840538},{"x":35.70898097753525,"y":81.59145677089691},{"x":39.94155639410019,"y":83.26475405693054},{"x":42.95827788114548,"y":86.43353658914566},{"x":39.764176189899445,"y":87.32476645708084},{"x":36.219153583049774,"y":85.95869213342667},{"x":64.10197567939758,"y":89.49556505680084},{"x":68.22205686569214,"y":86.13673758506775},{"x":71.8127316236496,"y":86.98408073186874},{"x":73.26841568946838,"y":89.27346312999725},{"x":71.92459321022034,"y":92.17298924922943},{"x":62.44171500205994,"y":93.04573714733124},{"x":31.96991491317749,"y":107.6513100862503},{"x":36.91116398572922,"y":108.80632710456848},{"x":37.83100974559784,"y":107.84696221351624},{"x":39.75290495157242,"y":108.15859878063202},{"x":41.08729958534241,"y":109.05613076686859},{"x":45.37723106145859,"y":114.44351887702942},{"x":49.86580538749695,"y":115.59975600242615},{"x":43.856497406959534,"y":108.97726941108704},{"x":41.35492968559265,"y":110.25707685947418},{"x":40.04151773452759,"y":111.69485104084015},{"x":36.815838396549225,"y":109.72805964946747},{"x":36.58897018432617,"y":108.29451870918274},{"x":32.02847445011139,"y":108.27437102794647},{"x":39.47241109609604,"y":111.86161315441132},{"x":40.09113299846649,"y":111.6380364894867},{"x":43.23629933595657,"y":113.00234961509705},{"x":49.69379901885986,"y":116.33892869949341},{"x":41.574267983436584,"y":107.91563832759857},{"x":39.02515745162964,"y":108.83277690410614},{"x":38.635754346847534,"y":107.74428224563599}]
\ No newline at end of file

test/data/facesFaceDescriptors.json

-[[-0.11301642656326294, 0.05311831831932068, 0.11299516260623932, 0.06697624176740646, -0.21439769864082336, 0.018440978601574898, -0.02246944047510624, -0.07858379185199738, 0.1299576610326767, -0.07627607882022858, 0.13028405606746674, -0.07381529361009598, -0.2666480243206024, 0.1386503130197525, -0.0318572111427784, 0.19745664298534393, -0.06815779209136963, -0.10036905854940414, -0.14179939031600952, -0.08129233121871948, 0.022137288004159927, 0.040264301002025604, 0.040526989847421646, 0.13235996663570404, -0.18518206477165222, -0.32066434621810913, -0.02661932446062565, -0.05834241956472397, 0.03743532672524452, -0.00874195247888565, -0.07306552678346634, 0.08665299415588379, -0.16994774341583252, -0.013914648443460464, 0.045591093599796295, 0.011259820312261581, -0.015375563874840736, -0.02478434517979622, 0.22394348680973053, 0.019354572519659996, -0.29177573323249817, 0.007424432784318924, 0.09004595875740051, 0.196985125541687, 0.17434047162532806, 0.07935666292905807, 0.10738393664360046, -0.08182166516780853, 0.09036535769701004, -0.35426175594329834, 0.004522169940173626, 0.14777712523937225, 0.03345843032002449, 0.019831661134958267, 0.05174245312809944, -0.21797265112400055, -0.0635240375995636, 0.1138913482427597, -0.14936882257461548, 0.13345786929130554, 0.09437544643878937, 0.013876287266612053, -0.05527258664369583, -0.09154847264289856, 0.19714784622192383, 0.12585902214050293, -0.1258784979581833, -0.14687272906303406, 0.15844950079917908, -0.22417423129081726, -0.009384687058627605, 0.16835945844650269, -0.0978832021355629, -0.23462620377540588, -0.2714555859565735, 0.013164130970835686, 0.5018162727355957, 0.15732169151306152, -0.17715050280094147, -0.05510217696428299, -0.028521616011857986, 0.016591474413871765, 0.06292184442281723, 0.06307955831289291, -0.05202113837003708, -0.04303480684757233, -0.06258511543273926, 0.04492054134607315, 0.2579847276210785, -0.030089672654867172, 0.034749776124954224, 0.2730059027671814, 0.10963069647550583, -0.018608057871460915, -0.03264498710632324, 0.17454616725444794, -0.14471100270748138, -0.05291951075196266, -0.15773987770080566, -0.00878031924366951, 0.013892430812120438, 0.03880078345537186, 0.06900060176849365, 0.07285363972187042, -0.1963559240102768, 0.1951112151145935, -0.1176108717918396, -0.0644276961684227, -0.05753452330827713, 0.0583438016474247, -0.13664516806602478, 0.008200451731681824, 0.18180623650550842, -0.19785352051258087, 0.08596935868263245, 0.18097352981567383, -0.02679232507944107, 0.0758017972111702, 0.04083810746669769, 0.06573939323425293, 0.028857335448265076, 0.02794247679412365, -0.1294119507074356, -0.12546013295650482, 0.09301090985536575, -0.18581253290176392, -0.00008967332541942596, 0.004743552301079035], [-0.09994355589151382, 0.10224375128746033, 0.05058574303984642, -0.007721252739429474, -0.017086468636989594, 0.07549440860748291, -0.035992272198200226, -0.05198703333735466, 0.21361662447452545, -0.12978722155094147, 0.240940660238266, 0.006456747651100159, -0.27226075530052185, -0.013018697500228882, -0.08444526791572571, 0.08976947516202927, -0.09226515889167786, -0.06434378027915955, -0.07935208082199097, -0.01867634989321232, 0.10694169998168945, 0.07422233372926712, 0.04553782194852829, -0.010077288374304771, -0.11393612623214722, -0.4091143012046814, -0.14219625294208527, -0.0598783977329731, -0.026971451938152313, -0.03787655383348465, 0.021513912826776505, -0.012842817232012749, -0.1919005811214447, -0.000048376619815826416, 0.014531563967466354, 0.06467536091804504, -0.07482472062110901, 0.012392744421958923, 0.1304527223110199, 0.014805854298174381, -0.13836951553821564, 0.02878289297223091, 0.033762313425540924, 0.28061288595199585, 0.10587205737829208, 0.0318145789206028, 0.00440259650349617, -0.11348624527454376, 0.036986637860536575, -0.26385852694511414, 0.1499217301607132, 0.14649057388305664, 0.11839955300092697, 0.10325036942958832, 0.17655009031295776, -0.09731947630643845, -0.03258107975125313, 0.14808624982833862, -0.15074148774147034, 0.10213347524404526, 0.06218596547842026, 0.06304526329040527, -0.019552389159798622, -0.10384533554315567, 0.13429109752178192, 0.14194521307945251, -0.16596026718616486, -0.19037191569805145, 0.04928822070360184, -0.1992926299571991, 0.017874792218208313, 0.1040906235575676, -0.08356981724500656, -0.2674674391746521, -0.26023608446121216, 0.12374836206436157, 0.42628446221351624, 0.20582066476345062, -0.14651867747306824, -0.026605937629938126, -0.057607077062129974, -0.06105601787567139, 0.06685034930706024, 0.00046340934932231903, -0.08286283910274506, 0.025767460465431213, -0.06461362540721893, 0.0723767876625061, 0.17591258883476257, -0.006310688331723213, -0.07715718448162079, 0.25646382570266724, -0.04985096678137779, 0.0062642693519592285, 0.00570443831384182, 0.046732913702726364, -0.10989770293235779, -0.003452849108725786, -0.1805979162454605, 0.024943644180893898, -0.07201632112264633, -0.10125844925642014, -0.04925733804702759, 0.021053487434983253, -0.13046781718730927, 0.1533353179693222, -0.01654619723558426, -0.06076410412788391, -0.014542249031364918, 0.05473938211798668, -0.26792800426483154, -0.009533338248729706, 0.1733490526676178, -0.29974472522735596, 0.20299524068832397, 0.1027199774980545, 0.08181609213352203, 0.16548770666122437, 0.1160709485411644, 0.08434373885393143, 0.01546089630573988, -0.02321481704711914, -0.11887650936841965, -0.07879133522510529, 0.04050300642848015, -0.018666662275791168, 0.02767147496342659, -0.0409514494240284], [-0.12437328696250916, 0.16371744871139526, 0.057088494300842285, -0.08345270156860352, -0.1361895501613617, -0.007239761762320995, -0.0027456339448690414, -0.0022191042080521584, 0.1767570823431015, -0.029083389788866043, 0.1353265941143036, -0.020310351625084877, -0.287823885679245, 0.03621964901685715, -0.1308278739452362, 0.1081429049372673, -0.23993228375911713, -0.10986220091581345, -0.05526617169380188, -0.16110815107822418, 0.0338473841547966, 0.1609116494655609, -0.031369976699352264, 0.04462018609046936, -0.12118063867092133, -0.26080891489982605, -0.005951732397079468, -0.1575908213853836, 0.0643172413110733, -0.1461418718099594, 0.07567547261714935, 0.107530377805233, -0.09414676576852798, -0.04453685134649277, 0.006743345409631729, 0.0339377298951149, -0.015679262578487396, -0.073453888297081, 0.21779049932956696, 0.05521396920084953, -0.1335650384426117, 0.010233767330646515, 0.0664336159825325, 0.430218368768692, 0.13693466782569885, -0.0009829048067331314, 0.0017257444560527802, -0.04437744617462158, 0.15366870164871216, -0.21540115773677826, 0.0167088583111763, 0.1910901665687561, 0.05646311119198799, 0.043141767382621765, 0.1210416778922081, -0.17660900950431824, 0.02089228481054306, 0.10505655407905579, -0.22187289595603943, 0.08246994763612747, 0.1212199255824089, 0.001875532791018486, 0.007675278931856155, -0.11065125465393066, 0.11531264334917068, -0.0005581285804510117, -0.08609281480312347, -0.12632550299167633, 0.11016186326742172, -0.18078474700450897, -0.1049041897058487, 0.1042267382144928, -0.10656270384788513, -0.11614145338535309, -0.22367611527442932, 0.02147454395890236, 0.3818414807319641, 0.19803355634212494, -0.04573599621653557, 0.03495173156261444, -0.02827727422118187, -0.07291625440120697, 0.09016091376543045, 0.022530820220708847, -0.16063527762889862, 0.008984813466668129, -0.06500966101884842, 0.06870897859334946, 0.1445840448141098, 0.07035897672176361, -0.09533844888210297, 0.16255708038806915, 0.03426520526409149, -0.07016268372535706, 0.020144084468483925, 0.0068038590252399445, -0.14746502041816711, 0.02669757604598999, -0.1343773752450943, -0.0007380247116088867, 0.036735646426677704, -0.16095727682113647, 0.003342229872941971, 0.09120983630418777, -0.18368342518806458, 0.12250301241874695, 0.014707496389746666, -0.011677264235913754, -0.09200786054134369, 0.054589901119470596, -0.17861320078372955, 0.08242300897836685, 0.2472871094942093, -0.2322787046432495, 0.17269502580165863, 0.16788843274116516, 0.047502290457487106, 0.1130535900592804, 0.10760728269815445, -0.010811767540872097, -0.047366999089717865, -0.04577387124300003, -0.12238828837871552, -0.1456669420003891, 0.06693950295448303, -0.025252800434827805, 0.12544143199920654, 0.0265400018543005], [0.025927569717168808, 0.15663708746433258, 0.058965474367141724, -0.028551187366247177, -0.12631410360336304, 0.022237226366996765, 0.01126374676823616, -0.1872863471508026, 0.1682029664516449, -0.002070157788693905, 0.15346597135066986, -0.18673506379127502, -0.278525710105896, 0.04948587715625763, -0.03619107976555824, 0.1413515955209732, -0.060395412147045135, -0.2583189606666565, -0.13808242976665497, -0.07600552588701248, 0.04279348626732826, -0.0666167140007019, 0.05459671467542648, 0.149562269449234, -0.23997187614440918, -0.3192921280860901, -0.06438620388507843, -0.06955230236053467, -0.016590990126132965, -0.13756340742111206, 0.016080018132925034, 0.10797879099845886, -0.1965179443359375, 0.0034376978874206543, 0.06646829843521118, 0.05213041231036186, -0.032798610627651215, -0.20469960570335388, 0.20994003117084503, 0.004714690148830414, -0.24007561802864075, -0.025683866813778877, 0.092221200466156, 0.29108813405036926, 0.21951010823249817, -0.012189175933599472, 0.11264907568693161, -0.060570698231458664, 0.03355713561177254, -0.25495976209640503, 0.05203044414520264, 0.2157290279865265, 0.04515017941594124, 0.07426977157592773, 0.045807648450136185, -0.12190262973308563, -0.033570460975170135, 0.16843782365322113, -0.1834886521100998, 0.09117748588323593, 0.15288689732551575, -0.06749839335680008, -0.08566926419734955, -0.19639848172664642, 0.2661381959915161, 0.18407370150089264, -0.1834886372089386, -0.16875605285167694, 0.12717457115650177, -0.18597885966300964, -0.11863835155963898, 0.05518629774451256, -0.18559770286083221, -0.16873560845851898, -0.28400951623916626, -0.014187514781951904, 0.45283856987953186, 0.17721779644489288, -0.2146177440881729, 0.011182520538568497, -0.09624088555574417, 0.10169383138418198, -0.035505618900060654, 0.1689070165157318, -0.0015639550983905792, 0.018879884853959084, -0.07630802690982819, 0.05508717522025108, 0.22385680675506592, -0.03961780667304993, -0.07674316316843033, 0.23865419626235962, -0.04872302711009979, -0.09690545499324799, -0.06860771775245667, 0.05698356404900551, -0.14728271961212158, 0.010753602720797062, -0.08920906484127045, -0.04629332199692726, -0.038761310279369354, 0.09933528304100037, 0.031854718923568726, 0.14741075038909912, -0.17927861213684082, 0.1629253327846527, -0.04969882220029831, 0.00284720491617918, -0.010596400126814842, -0.0009414497762918472, -0.046387046575546265, -0.0606590211391449, 0.14352722465991974, -0.3583222031593323, 0.06242058053612709, 0.1551140546798706, 0.052389003336429596, 0.04141424223780632, 0.01068437471985817, 0.10475203394889832, 0.11781509220600128, 0.034557536244392395, -0.16312718391418457, -0.06071166321635246, 0.024744128808379173, -0.029767779633402824, -0.00618806853890419, 0.03224271535873413], [-0.0962543934583664, -0.03510488569736481, -0.004949606955051422, 0.008284758776426315, -0.09557978063821793, -0.03148920461535454, -0.0035469243302941322, -0.03148280084133148, 0.15762394666671753, 0.018902912735939026, 0.20940224826335907, -0.0049227150157094, -0.23696963489055634, -0.006771348416805267, -0.053370341658592224, 0.06698818504810333, -0.13303788006305695, -0.08625766634941101, -0.08649979531764984, -0.10049654543399811, 0.06255536526441574, 0.11754642426967621, 0.02521955966949463, 0.08287493139505386, -0.13028612732887268, -0.29643961787223816, -0.11251382529735565, -0.1740148663520813, 0.02946341782808304, -0.14464221894741058, 0.003542322665452957, 0.15385982394218445, -0.14348679780960083, -0.08776889741420746, 0.06114386394619942, 0.044328175485134125, -0.028169168159365654, -0.019242623820900917, 0.1819373369216919, 0.12094853073358536, -0.11942562460899353, 0.013642464764416218, -0.00580492801964283, 0.4194585978984833, 0.17064036428928375, 0.03833707049489021, 0.023577354848384857, -0.0596885085105896, 0.10573002696037292, -0.3004778027534485, 0.13485471904277802, 0.162809818983078, 0.1528499722480774, 0.02323446422815323, 0.11265985667705536, -0.15413106977939606, 0.013475492596626282, 0.13343635201454163, -0.18546022474765778, 0.23613464832305908, 0.1127074807882309, -0.07481369376182556, -0.03090173937380314, -0.11048624664545059, 0.14306733012199402, 0.06088583171367645, -0.10998954623937607, -0.0625547245144844, 0.1647409200668335, -0.115761898458004, -0.07907391339540482, 0.11408128589391708, -0.17270313203334808, -0.2541918158531189, -0.265663743019104, 0.05659070611000061, 0.5032548904418945, 0.1484411507844925, -0.21916182339191437, -0.004813231527805328, -0.049820076674222946, -0.00413606408983469, 0.07484494149684906, 0.07051379978656769, -0.08695848286151886, -0.023705370724201202, -0.12480843812227249, 0.07716438174247742, 0.18726631999015808, 0.022125976160168648, -0.06583775579929352, 0.24017201364040375, -0.004352990537881851, -0.004003781825304031, 0.09011345356702805, 0.05776156485080719, -0.04413218796253204, -0.057881422340869904, -0.15213117003440857, -0.021110566332936287, -0.004055136814713478, -0.16205355525016785, -0.04073438048362732, 0.06350117921829224, -0.14136318862438202, 0.18425655364990234, 0.015558455139398575, -0.038499198853969574, -0.06519972532987595, -0.046765126287937164, -0.12038338929414749, 0.06919427961111069, 0.22361968457698822, -0.25389334559440613, 0.1741548627614975, 0.15434733033180237, 0.11286289244890213, 0.15635934472084045, -0.038396697491407394, 0.058131638914346695, -0.06190327927470207, -0.06887871026992798, -0.19518369436264038, -0.06716073304414749, 0.03750930726528168, -0.016213154420256615, -0.0019600093364715576, 0.03146345913410187], [-0.08438469469547272, 0.034433476626873016, 0.013390637934207916, -0.11032305657863617, -0.1721678525209427, -0.014739526435732841, 0.013159649446606636, -0.04383833706378937, 0.22945237159729004, -0.18055742979049683, 0.10724801570177078, -0.10486509650945663, -0.17328080534934998, 0.11445727944374084, -0.1428213119506836, 0.19179785251617432, -0.1129641979932785, -0.22180527448654175, -0.18176206946372986, -0.17263242602348328, 0.001636851578950882, 0.17328158020973206, -0.08476416766643524, 0.1016133576631546, -0.20712536573410034, -0.2548363208770752, -0.013701280578970909, -0.12087295949459076, 0.028648169711232185, -0.14270061254501343, 0.005408044904470444, 0.16749466955661774, -0.09364913403987885, 0.060463741421699524, 0.0917007252573967, 0.05608232319355011, -0.02542371302843094, -0.11898399889469147, 0.18762332201004028, 0.08870362490415573, -0.21150806546211243, -0.05339662358164787, 0.06960764527320862, 0.342587411403656, 0.24152815341949463, -0.055178068578243256, 0.006454486399888992, -0.12146948277950287, 0.1976870596408844, -0.29518774151802063, 0.03274869918823242, 0.12191275507211685, 0.030610322952270508, 0.14941972494125366, 0.09244397282600403, -0.18991874158382416, -0.0002718530595302582, 0.16477075219154358, -0.17079557478427887, 0.0982893779873848, 0.07169570028781891, -0.14121730625629425, -0.09334538877010345, -0.12710705399513245, 0.2545923888683319, 0.0946723222732544, -0.1666725128889084, -0.21636250615119934, 0.2331445813179016, -0.1981041133403778, -0.0874897837638855, 0.06129404529929161, -0.11788947135210037, -0.11157948523759842, -0.19723665714263916, 0.0467531718313694, 0.3331378996372223, 0.15040019154548645, -0.11666619032621384, 0.07890522480010986, -0.014858516864478588, 0.023824408650398254, -0.011289974674582481, 0.1570158302783966, -0.06161228567361832, -0.08446240425109863, -0.009107670746743679, 0.04418444260954857, 0.22349263727664948, -0.03572162613272667, -0.019860783591866493, 0.22091175615787506, -0.07901038229465485, -0.06882502138614655, -0.033531900495290756, 0.05211462080478668, -0.23550309240818024, 0.05773200839757919, -0.0063783712685108185, 0.00804441049695015, 0.0771300420165062, -0.01622505486011505, 0.031227124854922295, 0.15908673405647278, -0.12141232192516327, 0.24367079138755798, -0.15001045167446136, 0.03848479688167572, -0.008541088551282883, -0.055416084825992584, -0.03920590132474899, 0.006571357604116201, 0.17505057156085968, -0.2050420194864273, 0.187688946723938, 0.0819040909409523, 0.10374179482460022, 0.15251339972019196, 0.04850784316658974, 0.04916004836559296, 0.07320357859134674, 0.0493251234292984, -0.24662616848945618, -0.02634003758430481, 0.08344380557537079, -0.09143543243408203, -0.0650162473320961, 0.04578257352113724]]
\ No newline at end of file
+[[-0.11634846031665802,0.028185132890939713,0.13153618574142456,0.06227055937051773,-0.19837689399719238,-0.00583399273455143,-0.010891655460000038,-0.08271414786577225,0.11471451073884964,-0.08123917877674103,0.1272391676902771,-0.07800211757421494,-0.2540743947029114,0.13588032126426697,-0.0307321697473526,0.21247778832912445,-0.06502771377563477,-0.11616527289152145,-0.16744843125343323,-0.09554856270551682,0.016675475984811783,0.04842761904001236,0.0588490292429924,0.11036078631877899,-0.1645691692829132,-0.30832183361053467,-0.027835555374622345,-0.05220288783311844,0.06043601036071777,0.004156157374382019,-0.07474485784769058,0.08250502496957779,-0.17208999395370483,-0.037208594381809235,0.06834948062896729,0.024218283593654633,-0.010694803670048714,-0.024457599967718124,0.23895882070064545,0.0037566320970654488,-0.3064382076263428,0.030196093022823334,0.11086738109588623,0.19818036258220673,0.17517082393169403,0.10740934312343597,0.08606544882059097,-0.0939248651266098,0.06063482165336609,-0.3545883893966675,0.004837671294808388,0.13817209005355835,0.05888642370700836,0.028967585414648056,0.0347638800740242,-0.24132761359214783,-0.04302773252129555,0.11134087294340134,-0.16212642192840576,0.12925370037555695,0.0809628814458847,0.002201097086071968,-0.03398092836141586,-0.07904710620641708,0.19987981021404266,0.12266958504915237,-0.1274474859237671,-0.14003297686576843,0.15968921780586243,-0.22903524339199066,0.0004762057214975357,0.1669495701789856,-0.09815587848424911,-0.21806499361991882,-0.2793239653110504,-0.03180350363254547,0.4801402688026428,0.1383672058582306,-0.1833477020263672,-0.06399453431367874,-0.035138167440891266,0.03462325036525726,0.04942454770207405,0.061923690140247345,-0.07346022129058838,-0.040264375507831573,-0.05405578389763832,0.023277772590517998,0.2506774067878723,-0.039019107818603516,0.029081715270876884,0.27154070138931274,0.11957047879695892,-0.017708241939544678,-0.03387603536248207,0.1796264499425888,-0.13068023324012756,-0.059437572956085205,-0.163712278008461,-0.018396830186247826,0.005119161680340767,0.028815362602472305,0.08526614308357239,0.08816809207201004,-0.22821713984012604,0.18867996335029602,-0.12276707589626312,-0.06185315549373627,-0.046965308487415314,0.07672704011201859,-0.10262928903102875,-0.0063423877581954,0.18558722734451294,-0.19441832602024078,0.09239181876182556,0.2472042441368103,-0.03909676522016525,0.07437402755022049,0.04587577283382416,0.06336627155542374,0.022940605878829956,0.03857211768627167,-0.14419977366924286,-0.11545783281326294,0.07552200555801392,-0.18762795627117157,0.007402816787362099,0.03894437104463577],[-0.09233951568603516,0.07520502805709839,0.0471360944211483,-0.010466556996107101,0.003500264137983322,0.08081695437431335,-0.03387965261936188,-0.03733523562550545,0.21020442247390747,-0.09404481947422028,0.23934143781661987,0.0215359665453434,-0.2780716121196747,-0.005734823644161224,-0.09574554115533829,0.09024519473314285,-0.10668793320655823,-0.05615261569619179,-0.09537789970636368,-0.027305345982313156,0.09641921520233154,0.10733666270971298,0.05930484086275101,-0.013045689091086388,-0.09188339114189148,-0.4106191396713257,-0.13550889492034912,-0.08697408437728882,-0.010390311479568481,-0.02942235767841339,0.007921183481812477,-0.02016676962375641,-0.20198071002960205,-0.006702449172735214,0.0251946821808815,0.06029396504163742,-0.0691690444946289,0.0019519403576850891,0.13320721685886383,0.025177819654345512,-0.12530605494976044,0.035982027649879456,0.05226357653737068,0.2662340998649597,0.11323805153369904,0.04958421364426613,0.010302948765456676,-0.14813672006130219,0.028508305549621582,-0.26059791445732117,0.16988331079483032,0.1463342010974884,0.13083839416503906,0.10920721292495728,0.17165569961071014,-0.12292535603046417,-0.028080930933356285,0.13421514630317688,-0.14230471849441528,0.10813888162374496,0.062350884079933167,0.10053247213363647,-0.02928139455616474,-0.10359776765108109,0.1341533064842224,0.13900268077850342,-0.1618596464395523,-0.19058766961097717,0.04828851297497749,-0.20807933807373047,0.010564304888248444,0.10633311420679092,-0.08496777713298798,-0.2655319571495056,-0.2463349997997284,0.12063290178775787,0.4247370660305023,0.1920347362756729,-0.1615685671567917,-0.011293299496173859,-0.05202388018369675,-0.06393720954656601,0.08093788474798203,0.013875342905521393,-0.10013435781002045,0.036751288920640945,-0.058927882462739944,0.06755547970533371,0.17270849645137787,-0.0017291456460952759,-0.06124574691057205,0.25276318192481995,-0.04711493104696274,0.01840146631002426,0.02631673403084278,0.05701182782649994,-0.09648337215185165,0.008126601576805115,-0.18521596491336823,0.024786625057458878,-0.0613083615899086,-0.08818185329437256,-0.045461710542440414,0.009258095175027847,-0.1212489977478981,0.14221984148025513,-0.01942368969321251,-0.07132497429847717,0.010565785691142082,0.03708035871386528,-0.23937927186489105,-0.017209507524967194,0.18730974197387695,-0.28828105330467224,0.20588818192481995,0.09859722852706909,0.06930434703826904,0.16871735453605652,0.10860501974821091,0.12005928158760071,0.003807918168604374,-0.014734912663698196,-0.15592357516288757,-0.05718443542718887,0.04462910443544388,-0.025487978011369705,0.017820971086621284,-0.03943900391459465],[-0.11277016252279282,0.16098354756832123,0.05988795682787895,-0.11505985260009766,-0.13764601945877075,-0.018125953152775764,-0.014698371291160583,0.006905023939907551,0.1821925789117813,-0.021563978865742683,0.15091851353645325,-0.03721225634217262,-0.3001912534236908,0.016060858964920044,-0.12370184063911438,0.10668390244245529,-0.23569780588150024,-0.10505135357379913,-0.04109210893511772,-0.15012146532535553,0.04892900958657265,0.14137962460517883,-0.02200571820139885,0.0619843527674675,-0.09620106220245361,-0.26027771830558777,-0.01522919349372387,-0.15103775262832642,0.07414765655994415,-0.16027425229549408,0.06372426450252533,0.0926533192396164,-0.09527619183063507,-0.05952545255422592,-0.00499257817864418,0.053665850311517715,-0.02825658954679966,-0.07893257588148117,0.20068708062171936,0.03903854638338089,-0.1258913278579712,0.024329684674739838,0.04781285673379898,0.4228176474571228,0.15255069732666016,0.015188868157565594,0.00788198597729206,-0.03629109263420105,0.1480657160282135,-0.21167446672916412,0.013895329087972641,0.19699665904045105,0.07824041694402695,0.022653454914689064,0.12728801369667053,-0.16246333718299866,0.015902653336524963,0.1294557750225067,-0.2278294563293457,0.09121247380971909,0.11945690214633942,0.005948325619101524,0.010278221219778061,-0.12374573945999146,0.12210510671138763,0.009221900254487991,-0.09743078798055649,-0.1204695850610733,0.12769396603107452,-0.18469031155109406,-0.11325005441904068,0.10493327677249908,-0.11931980401277542,-0.10646456480026245,-0.2169141173362732,0.00949372723698616,0.3720436990261078,0.18269558250904083,-0.057609107345342636,0.030739817768335342,-0.015986662358045578,-0.05097445100545883,0.07677760720252991,0.021433746442198753,-0.16229106485843658,0.006307758390903473,-0.07988199591636658,0.05474869906902313,0.1320134699344635,0.05897970125079155,-0.08336386829614639,0.15999802947044373,0.04017291218042374,-0.054922763258218765,0.025332026183605194,-0.00094560906291008,-0.12491174787282944,0.023214157670736313,-0.1378774344921112,0.009755957871675491,0.021487023681402206,-0.18244768679141998,-0.01072671078145504,0.08217877149581909,-0.17091482877731323,0.11223600804805756,0.017901957035064697,-0.0034023327752947807,-0.10113242268562317,0.06516841799020767,-0.16639231145381927,0.07399401813745499,0.25069838762283325,-0.2423848956823349,0.17325596511363983,0.18680664896965027,0.05174680054187775,0.09361712634563446,0.08969344943761826,0.001638549380004406,-0.03986165672540665,-0.05375079810619354,-0.11656538397073746,-0.14891071617603302,0.04558948427438736,-0.030017226934432983,0.12282832711935043,0.03200910612940788],[0.021662116050720215,0.13764943182468414,0.05227668210864067,-0.04084555804729462,-0.136458158493042,0.04590526968240738,0.0188397578895092,-0.17823219299316406,0.18527886271476746,-0.017288826406002045,0.15326520800590515,-0.18968583643436432,-0.25598442554473877,0.060747843235731125,-0.04009188339114189,0.13465838134288788,-0.05551907420158386,-0.25349706411361694,-0.1359531283378601,-0.08201169967651367,0.03630130738019943,-0.07565895467996597,0.04985193908214569,0.14912472665309906,-0.24933066964149475,-0.2989639341831207,-0.07440325617790222,-0.07236435264348984,0.007793022319674492,-0.14277496933937073,0.015350904315710068,0.11266208440065384,-0.20027145743370056,-0.004338625818490982,0.07565931975841522,0.06843876838684082,-0.029365116730332375,-0.18628890812397003,0.2189212441444397,-0.000688113272190094,-0.24855753779411316,-0.009830232709646225,0.09292665123939514,0.29166179895401,0.2243579626083374,-0.019622473046183586,0.08338182419538498,-0.0414104089140892,0.028232108801603317,-0.25774481892585754,0.0500769168138504,0.21481895446777344,0.05254105478525162,0.06516268849372864,0.05102384835481644,-0.10823389887809753,-0.02942315861582756,0.19098074734210968,-0.17240288853645325,0.09920914471149445,0.1515028476715088,-0.08890038728713989,-0.07990352064371109,-0.17328815162181854,0.24634182453155518,0.1733505129814148,-0.17413198947906494,-0.1718866229057312,0.12582910060882568,-0.17642448842525482,-0.09674461930990219,0.07569882273674011,-0.18644993007183075,-0.17691926658153534,-0.271274596452713,-0.0020926538854837418,0.46468114852905273,0.19371949136257172,-0.2178988754749298,0.025562258437275887,-0.08117910474538803,0.11249618977308273,-0.018396683037281036,0.16247515380382538,0.0001282915472984314,0.018031783401966095,-0.0700790137052536,0.056407809257507324,0.21364818513393402,-0.036533765494823456,-0.07337557524442673,0.23493605852127075,-0.05353518947958946,-0.09696853905916214,-0.06235939636826515,0.04659310728311539,-0.1344364881515503,-0.012207120656967163,-0.11146484315395355,-0.06294745206832886,-0.034940220415592194,0.10326141119003296,0.03179185464978218,0.1467847377061844,-0.19082942605018616,0.1522996425628662,-0.05432801693677902,-0.01577235758304596,-0.022400803864002228,-0.000452450942248106,-0.036078453063964844,-0.040513791143894196,0.15848314762115479,-0.36531639099121094,0.05641081556677818,0.16581213474273682,0.03930053859949112,0.039935994893312454,0.033740460872650146,0.08137931674718857,0.13187022507190704,0.0346924290060997,-0.14783744513988495,-0.06158893555402756,0.02527773752808571,-0.04104948788881302,0.008458022028207779,0.015803828835487366],[-0.10051324218511581,-0.07279844582080841,-0.001282472163438797,0.020856289193034172,-0.06919786334037781,-0.024345869198441505,-0.010619491338729858,-0.013137751258909702,0.15405389666557312,0.016517357900738716,0.23002102971076965,-0.015056679956614971,-0.22843948006629944,-0.016521401703357697,-0.0438958965241909,0.1028967946767807,-0.11149217933416367,-0.08401092886924744,-0.08860547840595245,-0.07365550100803375,0.015243172645568848,0.1272270232439041,-0.016565930098295212,0.09345664083957672,-0.10889355093240738,-0.2854938209056854,-0.08980125188827515,-0.17001305520534515,0.040808405727148056,-0.15409760177135468,-0.035941049456596375,0.17426718771457672,-0.11924068629741669,-0.09771641343832016,0.09447836130857468,0.05560728535056114,-0.02027113363146782,-0.019314222037792206,0.22869902849197388,0.10169412195682526,-0.1377193033695221,-0.024289095774292946,0.011040525510907173,0.39462143182754517,0.1453857123851776,0.009099407121539116,0.015952548012137413,-0.060002267360687256,0.1221776157617569,-0.290526419878006,0.11363663524389267,0.15084052085876465,0.1706993579864502,0.04950060695409775,0.07974418997764587,-0.19215790927410126,-0.013606296852231026,0.141824871301651,-0.19528064131736755,0.21193517744541168,0.0698196291923523,-0.09058252722024918,-0.03922167792916298,-0.10907114297151566,0.1396980583667755,0.041026197373867035,-0.12431930005550385,-0.06791575998067856,0.17491401731967926,-0.0957411453127861,-0.03721051663160324,0.11379620432853699,-0.15487349033355713,-0.2650289833545685,-0.2548701763153076,0.054208435118198395,0.42744243144989014,0.11221148818731308,-0.2419632524251938,0.018017631024122238,-0.01796664670109749,-0.05460192635655403,0.042545948177576065,0.08495937287807465,-0.06071644648909569,0.007114706560969353,-0.07861480861902237,0.05576465278863907,0.1764492690563202,0.03052772767841816,-0.008522302843630314,0.21389426290988922,-0.024614224210381508,-0.0032321978360414505,0.10226888954639435,0.09135359525680542,-0.03673003613948822,-0.06905606389045715,-0.14689134061336517,-0.005894586443901062,0.002682933583855629,-0.15733204782009125,-0.007744046859443188,0.05100136250257492,-0.14602890610694885,0.20678772032260895,0.019124047830700874,-0.033549193292856216,-0.05248647928237915,-0.058607012033462524,-0.08463816344738007,0.06208709627389908,0.24236707389354706,-0.23599296808242798,0.17222557961940765,0.11931703239679337,0.09810569137334824,0.17590343952178955,-0.04135902225971222,0.06896808743476868,-0.06859633326530457,-0.06886428594589233,-0.2056688368320465,-0.061193596571683884,0.05314784497022629,-0.03325532376766205,0.002381695434451103,0.028421763330698013],[-0.07063059508800507,0.012398346327245235,-0.004086121916770935,-0.11135143041610718,-0.18017756938934326,-0.007133690640330315,0.008719323202967644,-0.051081277430057526,0.20928916335105896,-0.17685115337371826,0.10001912713050842,-0.10795890539884567,-0.1953105628490448,0.10430538654327393,-0.14811086654663086,0.18155619502067566,-0.11601020395755768,-0.21843558549880981,-0.19842198491096497,-0.15676547586917877,0.0026221834123134613,0.1657426804304123,-0.059600405395030975,0.11585110425949097,-0.18960151076316833,-0.2675485610961914,-0.0053065381944179535,-0.12873941659927368,0.021590830758213997,-0.14794206619262695,0.004776075482368469,0.17095273733139038,-0.09962760657072067,0.06954251229763031,0.08001348376274109,0.05619659274816513,-0.016173699870705605,-0.12300732731819153,0.20254217088222504,0.08233268558979034,-0.21961311995983124,-0.048164475709199905,0.060059789568185806,0.32081910967826843,0.24595360457897186,-0.05270609259605408,0.009464621543884277,-0.12156415730714798,0.19719916582107544,-0.3005487322807312,0.039065174758434296,0.1353994607925415,0.03396362066268921,0.13116101920604706,0.08811722695827484,-0.1892748475074768,0.009918665513396263,0.16392973065376282,-0.17915968596935272,0.0866990014910698,0.07109914720058441,-0.13285279273986816,-0.08271284401416779,-0.1261787712574005,0.26777613162994385,0.08791659772396088,-0.17506252229213715,-0.21035519242286682,0.2307121753692627,-0.2084561437368393,-0.08969969302415848,0.05810422822833061,-0.124051533639431,-0.11081314086914062,-0.20859161019325256,0.04635366052389145,0.33170226216316223,0.15251074731349945,-0.12201156467199326,0.086480051279068,-0.032260552048683167,0.015889232978224754,-0.021775195375084877,0.15573810040950775,-0.06982384622097015,-0.0784318596124649,-0.017058737576007843,0.04233776032924652,0.21266897022724152,-0.02595417946577072,-0.022498363628983498,0.2330653816461563,-0.06956781446933746,-0.07128044962882996,-0.016569793224334717,0.05142243206501007,-0.22927971184253693,0.047855496406555176,-0.007895056158304214,0.028033968061208725,0.06995373964309692,-0.02436123788356781,0.030047452077269554,0.1494320034980774,-0.11787243187427521,0.2352072298526764,-0.1698865294456482,0.02632801979780197,-0.007137699984014034,-0.07222921401262283,-0.05340973287820816,0.004369648173451424,0.17783141136169434,-0.22402727603912354,0.1835222989320755,0.07480072975158691,0.1268501579761505,0.15066304802894592,0.05315959453582764,0.05652406066656113,0.06715844571590424,0.05193927139043808,-0.2589290142059326,-0.031068526208400726,0.10136425495147705,-0.09585824608802795,-0.05503473058342934,0.05280270799994469]]
\ No newline at end of file

test/data/facesFaceLandmarkPositions.json

-[[{"x":78.70259515941143,"y":52.51417603031282},{"x":77.29212034493685,"y":67.79463638797883},{"x":75.96786513924599,"y":81.19702972904328},{"x":77.55778755992651,"y":95.10474457279328},{"x":82.7516676671803,"y":107.26841797367219},{"x":90.95986064523458,"y":118.21680511966828},{"x":102.14005157351494,"y":128.21510376468783},{"x":114.62478905916214,"y":135.2918554736722},{"x":127.32324719429016,"y":138.03325714603548},{"x":138.40828508138657,"y":134.74910034671908},{"x":146.35919749736786,"y":126.07305206791047},{"x":152.45407491922379,"y":117.23459686771515},{"x":158.06465715169907,"y":108.20900024906281},{"x":162.75949746370316,"y":97.26863159671906},{"x":166.2566688656807,"y":88.79356445804719},{"x":168.04164320230484,"y":79.34612335697297},{"x":169.36942607164383,"y":70.26872314945344},{"x":102.91276805102825,"y":50.643921513711255},{"x":111.27321466803551,"y":49.04517044559602},{"x":120.5909226834774,"y":48.130798955117505},{"x":130.22007316350937,"y":49.942017216836255},{"x":137.50274896621704,"y":54.37205566898469},{"x":151.10849559307098,"y":56.971192021523755},{"x":157.206868827343,"y":54.84915985599641},{"x":163.5188040137291,"y":54.879048009072584},{"x":168.17984849214554,"y":56.56761230960969},{"x":170.71705728769302,"y":59.195000310097974},{"x":144.00014102458954,"y":65.82685913578156},{"x":143.64073395729065,"y":73.73413528934601},{"x":144.01135712862015,"y":80.87621368900422},{"x":143.81918132305145,"y":89.1312643481839},{"x":127.48411267995834,"y":91.08588470951203},{"x":134.3283948302269,"y":93.52732910648469},{"x":139.52339559793472,"y":96.25963654056672},{"x":144.6911182999611,"y":94.77147736087922},{"x":148.38068902492523,"y":93.12768616214875},{"x":110.77253341674805,"y":61.845120091592115},{"x":116.32553532719612,"y":61.369480748330396},{"x":121.9557535648346,"y":62.41514648929719},{"x":126.42488598823547,"y":64.1528707934964},{"x":121.1322371661663,"y":64.05634369388703},{"x":115.86301743984222,"y":63.73924698368196},{"x":151.31528049707413,"y":67.09088959232453},{"x":156.5901842713356,"y":67.58717407718781},{"x":160.71268141269684,"y":67.98275055423859},{"x":164.26825791597366,"y":69.03151001468781},{"x":159.96803760528564,"y":70.4249998523343},{"x":155.32326072454453,"y":69.34288086429719},{"x":109.29639756679535,"y":99.49584259525422},{"x":120.37595704197884,"y":98.69051422611359},{"x":130.81436574459076,"y":99.85762085453156},{"x":135.84903299808502,"y":101.38172974125031},{"x":141.50265127420425,"y":100.36422028080109},{"x":146.36948823928833,"y":100.81127228275422},{"x":148.87848913669586,"y":103.27265228763703},{"x":145.39085745811462,"y":111.92646851078156},{"x":139.0690517425537,"y":116.24958862796906},{"x":133.29149305820465,"y":117.21682800785187},{"x":126.7722275853157,"y":115.69017853275422},{"x":117.6860836148262,"y":110.5495001269925},{"x":112.19027310609818,"y":100.30053009525422},{"x":128.99872034788132,"y":102.20095123783234},{"x":135.47163724899292,"y":103.91561188236359},{"x":140.5216109752655,"y":104.06739868656281},{"x":147.77944058179855,"y":103.93721832767609},{"x":139.77415084838867,"y":110.27076782718781},{"x":134.04826045036316,"y":111.13681854740265},{"x":129.13511961698532,"y":109.6996466113675}],[{"x":234.0432442203164,"y":72.18237426973158},{"x":234.07302695512772,"y":85.62781944967085},{"x":234.37448344379663,"y":97.4441898176747},{"x":237.17955499887466,"y":109.70111182428175},{"x":241.91123768687248,"y":120.36310103631789},{"x":248.58186745643616,"y":130.7120940992909},{"x":256.8855963945389,"y":140.60736063218886},{"x":265.9586812257767,"y":147.82223680711562},{"x":276.25926327705383,"y":150.3243668387013},{"x":285.8509864807129,"y":148.31700566507155},{"x":294.7637438774109,"y":140.88661578393751},{"x":301.7397997379303,"y":133.12361434198195},{"x":309.3003182411194,"y":123.77570560670668},{"x":315.0203056335449,"y":113.1747544119435},{"x":318.7541444301605,"y":102.62603953576857},{"x":320.53608107566833,"y":91.02094796396071},{"x":321.9833137989044,"y":79.67927816606337},{"x":243.00755888223648,"y":73.02197197175795},{"x":249.12357711791992,"y":70.65327099061781},{"x":256.6056842803955,"y":71.09805408216292},{"x":264.78340232372284,"y":72.58571747041518},{"x":271.8189814090729,"y":76.387691766985},{"x":287.77085518836975,"y":77.74872616029555},{"x":296.0825273990631,"y":74.829839498766},{"x":304.0993046760559,"y":74.12636736131483},{"x":311.1268253326416,"y":75.38659361100966},{"x":315.988929271698,"y":78.4906996796208},{"x":280.03022742271423,"y":84.97809437013441},{"x":279.16452836990356,"y":94.6116987774449},{"x":279.08406925201416,"y":103.45783117509657},{"x":278.3811960220337,"y":112.86081102586562},{"x":266.2560445070267,"y":112.55670645929152},{"x":271.9997515678406,"y":115.64479592538649},{"x":277.2981472015381,"y":118.46189263559157},{"x":282.88522934913635,"y":116.32307294107252},{"x":287.3393099308014,"y":114.4629829237538},{"x":250.71771156787872,"y":82.93822291589552},{"x":255.54227709770203,"y":81.77080169416243},{"x":261.47561633586884,"y":82.18437436319167},{"x":266.8718521595001,"y":85.09909179902846},{"x":261.02278220653534,"y":85.3523221323567},{"x":255.04803681373596,"y":85.51701000428969},{"x":289.57169246673584,"y":86.59227958417708},{"x":296.12601137161255,"y":85.3894476006108},{"x":301.7222509384155,"y":85.33728733754927},{"x":306.4811248779297,"y":87.00033477044875},{"x":301.593781709671,"y":89.3077004263478},{"x":295.20851707458496,"y":88.81555894113356},{"x":255.08908998966217,"y":120.65163496232802},{"x":262.8024332523346,"y":120.9952745745259},{"x":270.70326828956604,"y":122.40383675790602},{"x":275.5565803050995,"y":124.17473438478285},{"x":281.3999750614166,"y":122.6351008722859},{"x":288.9069309234619,"y":122.48449710107619},{"x":295.68931913375854,"y":123.3941626856404},{"x":289.19691920280457,"y":131.8845453569966},{"x":280.730872631073,"y":135.16087916589552},{"x":274.73176169395447,"y":135.9497754881459},{"x":269.14574098587036,"y":134.31138041711623},{"x":261.4882484674454,"y":129.7903356859761},{"x":257.0808787345886,"y":121.76479747987563},{"x":269.8720760345459,"y":124.05334284997755},{"x":275.90254950523376,"y":125.26417449212843},{"x":281.2652223110199,"y":125.28600934243971},{"x":293.8986005783081,"y":124.10085681176955},{"x":280.6464831829071,"y":130.53398588395888},{"x":274.9977023601532,"y":131.01732567048842},{"x":270.30089259147644,"y":129.57399228311354}],[{"x":53.439819291234016,"y":280.0604960024357},{"x":51.78455471992493,"y":295.82864955067635},{"x":50.52304560318589,"y":309.732116997242},{"x":51.996403969824314,"y":323.6099224090576},{"x":55.675161108374596,"y":336.0948809981346},{"x":61.96380540728569,"y":348.32347297668457},{"x":70.6162890791893,"y":359.93272376060486},{"x":80.40928453207016,"y":369.4880121946335},{"x":91.329245865345,"y":373.68778198957443},{"x":102.01146513223648,"y":372.54405105113983},{"x":111.70319825410843,"y":365.45324832201004},{"x":120.67031770944595,"y":358.51266556978226},{"x":130.24644166231155,"y":350.4673351049423},{"x":138.3169224858284,"y":340.3098750114441},{"x":144.8827138543129,"y":330.19956386089325},{"x":149.44279700517654,"y":318.41588428616524},{"x":153.5925915837288,"y":306.7800603210926},{"x":65.65944314002991,"y":287.4393824636936},{"x":74.17677998542786,"y":286.2525469958782},{"x":83.42234998941422,"y":288.1453210115433},{"x":92.45077580213547,"y":291.6930521428585},{"x":100.51797538995743,"y":297.3068634867668},{"x":118.6282268166542,"y":301.29853051900864},{"x":127.82300025224686,"y":299.62827998399734},{"x":136.26429945230484,"y":299.9250599741936},{"x":144.1239109635353,"y":301.3865695595741},{"x":148.95386010408401,"y":305.0852126777172},{"x":108.77183586359024,"y":309.15458089113235},{"x":106.31233364343643,"y":318.45475602149963},{"x":105.28086930513382,"y":327.17767184972763},{"x":103.4764102101326,"y":336.8808351159096},{"x":89.56375032663345,"y":334.2973544597626},{"x":95.11743992567062,"y":338.2826396226883},{"x":100.4269888997078,"y":342.33478063344955},{"x":107.09357768297195,"y":340.92600989341736},{"x":111.45647674798965,"y":339.16702830791473},{"x":74.23168629407883,"y":299.72840997576714},{"x":80.10014206171036,"y":300.2405780553818},{"x":86.55913084745407,"y":301.92818266153336},{"x":92.06215351819992,"y":305.3237600326538},{"x":85.29800087213516,"y":304.93414691090584},{"x":78.56964975595474,"y":303.802222430706},{"x":118.96386057138443,"y":311.23431676626205},{"x":125.75672775506973,"y":311.30804657936096},{"x":131.75824791193008,"y":312.1569736003876},{"x":136.9495204091072,"y":313.76811495423317},{"x":130.84956914186478,"y":315.8736974000931},{"x":124.32669311761856,"y":314.29629921913147},{"x":73.89956772327423,"y":338.6452066898346},{"x":83.92988950014114,"y":341.8424386382103},{"x":92.97652751207352,"y":345.22819715738297},{"x":97.08906918764114,"y":347.40902799367905},{"x":103.12756806612015,"y":346.28637260198593},{"x":109.98074918985367,"y":346.54456919431686},{"x":116.1731830239296,"y":346.1622644662857},{"x":108.54592829942703,"y":355.2732748389244},{"x":100.48217684030533,"y":358.47334402799606},{"x":94.37009304761887,"y":358.60381096601486},{"x":88.47060412168503,"y":356.5071900486946},{"x":80.98365873098373,"y":350.6593140363693},{"x":77.15098470449448,"y":340.4296311736107},{"x":91.45873099565506,"y":346.6854547262192},{"x":97.10512071847916,"y":348.816522359848},{"x":102.38213449716568,"y":349.28899598121643},{"x":113.6016657948494,"y":347.39245051145554},{"x":101.5955975651741,"y":352.2493990659714},{"x":95.40374487638474,"y":352.52306592464447},{"x":91.18230849504471,"y":350.25800210237503}],[{"x":253.64455546438694,"y":274.96428267994236},{"x":255.63555194437504,"y":286.48764054813694},{"x":257.851626470685,"y":296.23062263765644},{"x":261.372583928518,"y":306.3042299036057},{"x":265.716976583004,"y":315.1606678847344},{"x":271.16881239414215,"y":323.31880401411365},{"x":278.9179730415344,"y":330.80829064884495},{"x":287.70828092098236,"y":335.8823236111672},{"x":297.32989490032196,"y":337.5626749877007},{"x":307.2697219848633,"y":337.14021955766987},{"x":316.2518972158432,"y":331.886222172168},{"x":323.5890635251999,"y":326.10143922605823},{"x":330.2825126647949,"y":317.92637084760975},{"x":335.62895596027374,"y":308.1591898683579},{"x":339.1303542852402,"y":298.46708123722385},{"x":341.2391484975815,"y":288.10848497190784},{"x":343.8476446866989,"y":277.8678704980881},{"x":257.7795196324587,"y":265.5684605125458},{"x":262.86759052053094,"y":259.46244253554653},{"x":271.30318570137024,"y":258.1320998553307},{"x":280.2511401772499,"y":259.1452731970818},{"x":288.77513176202774,"y":262.10564820804905},{"x":300.88161540031433,"y":262.23092802205394},{"x":311.1703190803528,"y":259.23369329014133},{"x":320.83708584308624,"y":259.57886569538425},{"x":329.8163455724716,"y":262.03275864162754},{"x":335.06606125831604,"y":268.59239073434185},{"x":295.74934458732605,"y":272.0405581776173},{"x":295.10603761672974,"y":281.74871687927555},{"x":294.78596317768097,"y":290.4453236345322},{"x":294.52324163913727,"y":299.96156142988514},{"x":283.94295251369476,"y":300.7173269275696},{"x":289.2285826802254,"y":303.9793581847222},{"x":294.77881556749344,"y":307.2272955063851},{"x":300.44683516025543,"y":304.92734794655155},{"x":305.3025699853897,"y":303.13202046194385},{"x":267.63354451954365,"y":273.45072929897617},{"x":271.67723220586777,"y":269.8547673110039},{"x":277.1182727217674,"y":269.4767764691384},{"x":282.23915058374405,"y":273.35298531093906},{"x":276.9609106183052,"y":274.3409555439026},{"x":271.4951269328594,"y":275.01823853292774},{"x":306.1736521720886,"y":273.4907934550316},{"x":311.9750416278839,"y":270.6327790502579},{"x":317.42995393276215,"y":270.95207189359974},{"x":321.920796751976,"y":274.85076408424686},{"x":317.8245964050293,"y":276.6930876258881},{"x":311.29924952983856,"y":276.1408682648213},{"x":279.1467461884022,"y":314.23833452501606},{"x":284.48106986284256,"y":313.8286061767609},{"x":290.1339142918587,"y":313.8490322712929},{"x":294.4958354830742,"y":315.5466157678635},{"x":299.6633150577545,"y":313.5960527543099},{"x":306.1217950582504,"y":314.4668222550423},{"x":312.643159031868,"y":316.36552153864216},{"x":306.774823307991,"y":320.4127126220734},{"x":300.28556072711945,"y":321.05349503794025},{"x":294.83206647634506,"y":321.65074037351917},{"x":289.9620077610016,"y":320.5292926911385},{"x":284.2809522151947,"y":318.41614382305454},{"x":280.899098277092,"y":314.7755777720482},{"x":289.82488214969635,"y":316.4206280592949},{"x":295.0254775881767,"y":317.5740782503159},{"x":299.5264849662781,"y":317.32291888037037},{"x":310.37555599212646,"y":316.52424393930744},{"x":299.7602632045746,"y":316.030502903846},{"x":294.2890642285347,"y":316.5608839873345},{"x":290.4010553956032,"y":315.42048167028736}],[{"x":458.6051552891731,"y":217.72117600517888},{"x":460.09398763626814,"y":230.39350793438572},{"x":461.8694317564368,"y":241.38930038290638},{"x":465.2278592949733,"y":252.93564961033482},{"x":469.7518924251199,"y":263.28325144129417},{"x":476.1563876569271,"y":272.5680155523362},{"x":485.33472272753716,"y":280.3438511856141},{"x":495.19925662875175,"y":285.54824075775764},{"x":505.49679857492447,"y":287.5497344501557},{"x":517.0163440704346,"y":286.52682790833137},{"x":528.832253575325,"y":281.7315962679925},{"x":538.1082131266594,"y":275.27146718578956},{"x":545.7302697896957,"y":266.1297963030877},{"x":550.5768427848816,"y":254.52256647425313},{"x":554.072232067585,"y":242.63067278938908},{"x":554.7430681586266,"y":229.65774825888295},{"x":555.9324762821198,"y":217.0304150350632},{"x":464.2269563060254,"y":212.73671243267674},{"x":469.75961059704423,"y":209.3460837967934},{"x":477.53361535072327,"y":208.5074197419228},{"x":485.2873308509588,"y":209.6281631358208},{"x":492.1645346879959,"y":212.7300988801064},{"x":509.10780626535416,"y":212.85144624786992},{"x":518.4267295002937,"y":209.07269889789242},{"x":527.5292375683784,"y":208.5795753963532},{"x":536.2015252709389,"y":209.3796590931954},{"x":542.8177468776703,"y":212.6919245489182},{"x":501.580209761858,"y":221.49086973744053},{"x":500.3731763958931,"y":230.08853492813725},{"x":500.1400006711483,"y":238.13611183243412},{"x":499.29766792058945,"y":247.00497171955723},{"x":491.3476411700249,"y":249.0014759667458},{"x":496.11430510878563,"y":251.5378865964951},{"x":500.55614736676216,"y":253.75804135876317},{"x":505.92355754971504,"y":252.01613936501164},{"x":511.2390627861023,"y":250.3659447796883},{"x":473.7797954529524,"y":223.44442883806843},{"x":477.8101897537708,"y":221.86633915382046},{"x":484.2290563881397,"y":221.8739053078236},{"x":489.7755623757839,"y":224.0043203301968},{"x":483.90396946668625,"y":225.28688494043965},{"x":478.22655925154686,"y":225.68062556462903},{"x":515.0914766192436,"y":223.67829639392514},{"x":520.7296285629272,"y":222.6548081405701},{"x":526.719929933548,"y":222.64345110135693},{"x":532.6925768852234,"y":223.93850103693623},{"x":527.5918874144554,"y":225.99143520670552},{"x":520.3674954771996,"y":225.9896002777161},{"x":487.79902869462967,"y":262.3062415488305},{"x":492.6807407736778,"y":261.1754646070542},{"x":498.5045930147171,"y":260.4150388010087},{"x":502.8779262602329,"y":262.09455320912025},{"x":507.8618753552437,"y":259.85850093441627},{"x":515.5100157856941,"y":260.53184179144523},{"x":522.529854118824,"y":263.0671280391755},{"x":516.0406350493431,"y":268.1145648129525},{"x":509.29703161120415,"y":269.5725134976449},{"x":503.7219538986683,"y":270.14516750651023},{"x":498.86963579058647,"y":268.82049713211677},{"x":493.03908091783524,"y":266.78845772820137},{"x":490.25995284318924,"y":262.70426497536323},{"x":498.5697175562382,"y":263.05846873598716},{"x":503.7999505698681,"y":264.6741432316842},{"x":508.1411578655243,"y":264.4359304078164},{"x":519.9255567789078,"y":263.10483387785575},{"x":508.362724006176,"y":264.35242161827705},{"x":502.9432258605957,"y":264.6728236444535},{"x":499.6765456497669,"y":263.62174508648536}],[{"x":575.6323959156871,"y":97.81703020103517},{"x":576.2381174787879,"y":110.7464268899733},{"x":577.1207508929074,"y":122.03858585126939},{"x":579.777023255825,"y":133.77679456241668},{"x":584.0378794074059,"y":144.10788560875},{"x":590.4641489386559,"y":153.66901643045486},{"x":599.8314836025238,"y":162.39560646780075},{"x":609.7557654380798,"y":168.6680013037497},{"x":620.3408677577972,"y":170.72851402528823},{"x":630.3204867839813,"y":168.76506347425521},{"x":639.1392222642899,"y":162.16524405248703},{"x":646.3353884220123,"y":155.17852164276184},{"x":652.9094681739807,"y":146.6903299189383},{"x":657.3339067697525,"y":136.69562412269653},{"x":660.4557708501816,"y":127.19812996156755},{"x":662.051419377327,"y":116.94922510750833},{"x":663.7066377401352,"y":106.84000776179376},{"x":586.5404289513826,"y":98.32077897556367},{"x":593.4320383369923,"y":96.36920688517633},{"x":601.3935153484344,"y":96.69368816383424},{"x":609.6317137479782,"y":98.59984148510041},{"x":616.5242365002632,"y":102.5144240654284},{"x":632.4123371839523,"y":103.04483480222764},{"x":640.0241969823837,"y":100.34534380801263},{"x":647.2620248794556,"y":99.55671287544313},{"x":654.129354596138,"y":100.25115479000154},{"x":659.0634310245514,"y":102.63225520618501},{"x":624.9623003005981,"y":109.84366706975045},{"x":624.2144320011139,"y":118.4606404400641},{"x":624.1961280107498,"y":126.39726979501786},{"x":623.8329563140869,"y":134.92106182344497},{"x":612.6639273762703,"y":135.8878824210936},{"x":617.6502031683922,"y":138.31433464057983},{"x":622.2424348592758,"y":140.77128500707687},{"x":627.0541898012161,"y":139.33917314775528},{"x":630.7270992994308,"y":137.66990662582458},{"x":595.4028062522411,"y":108.46945513256135},{"x":600.3049533367157,"y":107.04665152199807},{"x":606.116753757,"y":107.49907577999177},{"x":610.7790024876595,"y":110.7087984777266},{"x":605.489963054657,"y":111.36523569591584},{"x":599.8057773709297,"y":111.40001646645608},{"x":634.6385422945023,"y":112.09705091484132},{"x":640.4470851421356,"y":110.56770468719544},{"x":645.5006678104401,"y":110.40809006460252},{"x":650.1414160728455,"y":112.40711540945115},{"x":645.6796168088913,"y":114.66480738885942},{"x":639.6736890077591,"y":114.49185062416139},{"x":602.1105405688286,"y":143.5170047379309},{"x":610.0115154385567,"y":144.67639989622177},{"x":616.9459272027016,"y":145.61259479291977},{"x":620.7917128801346,"y":146.8697983122641},{"x":625.5570113658905,"y":145.48805750377716},{"x":631.3872159719467,"y":145.44158322580398},{"x":637.3369463682175,"y":145.60786254175247},{"x":631.6403584480286,"y":151.87812722690643},{"x":624.9396952390671,"y":154.0825080371672},{"x":620.1312247514725,"y":154.83688754804672},{"x":615.6493668556213,"y":153.58939088352264},{"x":608.7029096484184,"y":150.48222155340255},{"x":604.0386838912964,"y":144.4577150440985},{"x":616.3138701319695,"y":147.00049645669998},{"x":621.2553536891937,"y":148.18779403932632},{"x":625.1526666879654,"y":148.3215981221968},{"x":635.5966547727585,"y":146.52967573173584},{"x":624.8098908662796,"y":149.64450467594207},{"x":620.2795970439911,"y":149.9581789470488},{"x":616.9711862206459,"y":148.92591042287887}]]
\ No newline at end of file
+[[{"x":78.88453498482704,"y":54.340515751992505},{"x":78.80174972116947,"y":67.42203964725617},{"x":79.41133201122284,"y":79.55764832035187},{"x":80.87148245424032,"y":90.45394958988312},{"x":83.65095367655158,"y":103.08840240970734},{"x":90.00872641801834,"y":113.85682739750031},{"x":98.88908013701439,"y":122.59016479984406},{"x":110.19187986850739,"y":132.62888397709017},{"x":123.95810306072235,"y":138.49772705570345},{"x":136.95235967636108,"y":134.42995514408236},{"x":142.33598232269287,"y":126.9210821582425},{"x":147.8650164604187,"y":120.19905151859406},{"x":153.60961079597473,"y":110.54403748050812},{"x":159.75567817687988,"y":99.66677155033234},{"x":163.89759451150894,"y":89.78836121097687},{"x":166.517576277256,"y":80.12262024418},{"x":166.97096794843674,"y":69.13838638797883},{"x":101.14158697426319,"y":48.7663084460843},{"x":109.53669220209122,"y":45.81676353946809},{"x":117.78285771608353,"y":45.70639671817903},{"x":124.79132235050201,"y":47.50745834842805},{"x":130.46167314052582,"y":50.689716954385084},{"x":151.67625218629837,"y":53.574116368447584},{"x":156.42495572566986,"y":52.002991337930005},{"x":161.256083548069,"y":51.58568253055696},{"x":166.25041037797928,"y":52.67973007694368},{"x":168.02350252866745,"y":56.0030027820218},{"x":139.7132870554924,"y":65.73578323856476},{"x":140.8638072013855,"y":74.94289841190461},{"x":141.84472680091858,"y":83.16134705082062},{"x":141.3910087943077,"y":90.50144257084015},{"x":128.617801964283,"y":91.91793121830109},{"x":132.71707564592361,"y":94.04132904544953},{"x":137.27396100759506,"y":95.4687887622464},{"x":141.28654569387436,"y":94.48701538578156},{"x":143.8589584827423,"y":93.67591156498078},{"x":109.83383402228355,"y":59.16733994022493},{"x":115.58296576142311,"y":58.19477142826204},{"x":121.17200314998627,"y":58.62158646122102},{"x":125.14821708202362,"y":62.44804062381868},{"x":121.36037021875381,"y":63.82070984378938},{"x":114.78775784373283,"y":62.54481186405305},{"x":146.6951486468315,"y":65.14801468387726},{"x":152.43364095687866,"y":62.580219884072584},{"x":158.494673371315,"y":63.33990921512727},{"x":159.57544267177582,"y":65.69073929325226},{"x":156.95089668035507,"y":67.93409409061555},{"x":151.64972066879272,"y":66.9153219653714},{"x":110.81738442182541,"y":102.46230377689484},{"x":121.06627717614174,"y":100.86323608890656},{"x":131.50526702404022,"y":100.96113647953156},{"x":135.66662937402725,"y":101.99545349613312},{"x":139.55124586820602,"y":101.70082916751984},{"x":145.57576447725296,"y":104.24478210941437},{"x":147.33246833086014,"y":106.75124420658234},{"x":141.4960440993309,"y":114.78596558109406},{"x":136.66154563426971,"y":118.56308044925812},{"x":130.54882258176804,"y":119.1232305957425},{"x":124.78052198886871,"y":117.28310455814484},{"x":117.97604158520699,"y":112.96259369388703},{"x":111.7592890560627,"y":103.06800140873078},{"x":127.85126686096191,"y":103.7278219653714},{"x":134.48065489530563,"y":104.94294037357453},{"x":139.40277218818665,"y":105.46841492191437},{"x":145.6920874118805,"y":106.89988770023469},{"x":137.0137956738472,"y":113.12150635257844},{"x":131.9087651371956,"y":113.30772461429719},{"x":126.09670430421829,"y":111.97380127445344}],[{"x":230.35409008711576,"y":76.48713436818892},{"x":231.32800871878862,"y":88.98580232358748},{"x":232.9434676747769,"y":100.94655373788649},{"x":234.67413983494043,"y":110.8506565401631},{"x":237.67880927026272,"y":122.19467714525038},{"x":243.33399283885956,"y":131.75854900575453},{"x":250.60516518354416,"y":139.53664401269728},{"x":260.6413799524307,"y":148.7868421385365},{"x":275.1246545314789,"y":153.69684174753004},{"x":289.2145631313324,"y":149.69657686448866},{"x":297.723708152771,"y":141.7915036985951},{"x":304.31809639930725,"y":135.26991513467604},{"x":310.15464186668396,"y":125.56473901964003},{"x":314.48908019065857,"y":114.62088111139113},{"x":316.87436509132385,"y":104.21722486711317},{"x":319.2071588039398,"y":93.50209012723738},{"x":319.9787652492523,"y":81.70821490979964},{"x":241.29130882024765,"y":70.10276725984389},{"x":247.62181669473648,"y":67.91886738038832},{"x":254.63222289085388,"y":68.07646480298811},{"x":261.27166974544525,"y":69.60374274946028},{"x":267.1634376049042,"y":72.26310375428969},{"x":289.5264410972595,"y":73.65641334749037},{"x":295.1576931476593,"y":71.68943074441725},{"x":301.7056484222412,"y":70.84351006246382},{"x":308.6310956478119,"y":71.41338721013838},{"x":313.5512583255768,"y":73.57223716474348},{"x":277.74836778640747,"y":85.97259298063094},{"x":277.7395558357239,"y":95.77593258119398},{"x":277.6018691062927,"y":105.088952572115},{"x":277.0711143016815,"y":112.56092408395583},{"x":268.3519619703293,"y":112.92281702256972},{"x":271.9692051410675,"y":115.1969049284535},{"x":276.8920214176178,"y":116.6187560865956},{"x":281.53438234329224,"y":115.31924370027357},{"x":285.0949912071228,"y":113.99634078241164},{"x":249.88221472501755,"y":79.41814842916304},{"x":254.2900288105011,"y":78.42612508035475},{"x":260.09243297576904,"y":78.6968239853459},{"x":265.1237539052963,"y":82.23297038770491},{"x":260.5498933792114,"y":83.62710276342207},{"x":254.27618503570557,"y":82.63636890149886},{"x":289.2370285987854,"y":83.55092278219038},{"x":294.7169153690338,"y":80.5464356253224},{"x":300.86022186279297,"y":81.0058490345555},{"x":304.41960406303406,"y":83.01301661229903},{"x":300.4358162879944,"y":85.69828143811995},{"x":294.3155460357666,"y":85.09616366124922},{"x":255.21712112426758,"y":121.53793075776869},{"x":262.8897992372513,"y":121.6964166472035},{"x":272.0220493078232,"y":122.16711047387892},{"x":276.60698986053467,"y":123.12763646341139},{"x":281.10380840301514,"y":122.24365690446669},{"x":290.48466062545776,"y":123.39597419000441},{"x":297.0332818031311,"y":123.55629924035841},{"x":289.114342212677,"y":132.51426437593275},{"x":282.61101245880127,"y":136.3083288977223},{"x":275.5814690589905,"y":137.46408251024062},{"x":269.35472190380096,"y":136.106339246996},{"x":262.0876613855362,"y":131.79087260461623},{"x":256.09996592998505,"y":122.10302332139784},{"x":269.7251617908478,"y":124.58125117517287},{"x":276.5752167701721,"y":125.28405955529982},{"x":283.01117515563965,"y":125.34827116227919},{"x":295.4524829387665,"y":123.87910655237013},{"x":282.24452543258667,"y":131.15521434045607},{"x":275.948046207428,"y":132.09439638353163},{"x":269.59646224975586,"y":131.0175676653462}],[{"x":52.89241414517164,"y":282.7266910672188},{"x":50.87797347456217,"y":296.89296850562096},{"x":49.71260245889425,"y":310.34687554836273},{"x":49.50424597132951,"y":322.08700078725815},{"x":50.409748582169414,"y":336.39450591802597},{"x":55.15722617506981,"y":349.3232504725456},{"x":62.49335914850235,"y":360.313629090786},{"x":72.71078884601593,"y":371.9817545413971},{"x":87.83086746931076,"y":380.3500460386276},{"x":104.21213418245316,"y":378.6363449692726},{"x":114.80719953775406,"y":372.0393912792206},{"x":123.89140039682388,"y":366.18297278881073},{"x":132.48398214578629,"y":356.70495879650116},{"x":139.61599975824356,"y":345.4006921648979},{"x":144.25993472337723,"y":334.65290147066116},{"x":149.00599032640457,"y":323.61125260591507},{"x":151.8666437268257,"y":310.911751717329},{"x":67.48781859874725,"y":283.3429951071739},{"x":75.35115391016006,"y":282.75306625664234},{"x":83.2165738940239,"y":285.16011513769627},{"x":90.33202439546585,"y":289.1681001484394},{"x":96.4085391163826,"y":293.61568254232407},{"x":121.10100895166397,"y":299.9678646326065},{"x":127.67173558473587,"y":298.52197524905205},{"x":135.01537948846817,"y":298.0430044233799},{"x":142.36321359872818,"y":298.94229823350906},{"x":146.83973103761673,"y":301.76432624459267},{"x":105.43120414018631,"y":309.11801970005035},{"x":103.4981182217598,"y":318.78210747241974},{"x":101.45671278238297,"y":327.84984415769577},{"x":99.53793495893478,"y":335.4310358762741},{"x":89.49993461370468,"y":333.841296851635},{"x":93.19077640771866,"y":336.89287608861923},{"x":98.23556751012802,"y":339.2234809398651},{"x":103.67715865373611,"y":338.73299968242645},{"x":107.70168215036392,"y":337.7921968102455},{"x":74.69314396381378,"y":296.5739365518093},{"x":80.14858573675156,"y":297.56556755304337},{"x":86.56933814287186,"y":298.90682503581047},{"x":91.52818530797958,"y":303.0784338712692},{"x":86.24918073415756,"y":303.6382160782814},{"x":79.09003406763077,"y":301.4291281700134},{"x":118.30803662538528,"y":309.0543663203716},{"x":125.0142565369606,"y":307.85860392451286},{"x":131.8498483300209,"y":309.31628438830376},{"x":135.02473145723343,"y":311.04234552383423},{"x":130.3895464539528,"y":313.3754262328148},{"x":123.57992440462112,"y":311.814509421587},{"x":73.74992281198502,"y":341.166852414608},{"x":82.1018448472023,"y":342.58997851610184},{"x":91.98869973421097,"y":344.52026331424713},{"x":96.75153106451035,"y":346.0399171113968},{"x":101.20932966470718,"y":346.01455879211426},{"x":110.74226409196854,"y":348.3536369204521},{"x":117.45747834444046,"y":349.5986934900284},{"x":107.76521474123001,"y":356.0601131916046},{"x":100.55026322603226,"y":358.1020730137825},{"x":93.41589421033859,"y":357.8789767026901},{"x":87.21731215715408,"y":355.56834787130356},{"x":80.43258100748062,"y":350.9654130935669},{"x":74.80048149824142,"y":341.89357817173004},{"x":88.80335599184036,"y":346.7158262729645},{"x":95.74061542749405,"y":348.37989485263824},{"x":102.56206899881363,"y":349.5257633328438},{"x":115.78181892633438,"y":349.6036605834961},{"x":101.4380207657814,"y":352.6228076219559},{"x":95.14025956392288,"y":352.4510815143585},{"x":88.81180793046951,"y":350.6978974342346}],[{"x":253.42183826863766,"y":272.33237447181057},{"x":254.645193323493,"y":283.81189538874935},{"x":256.80082605034113,"y":294.49671291151355},{"x":259.1452670022845,"y":303.39569072761844},{"x":262.62326591834426,"y":313.8439553860695},{"x":268.03879207372665,"y":322.1283790353806},{"x":274.28707036376,"y":328.6552868370087},{"x":282.563962161541,"y":336.6665140871079},{"x":295.7026590704918,"y":341.3614055398972},{"x":309.2031053304672,"y":337.19024919309925},{"x":317.6761395931244,"y":329.91151605882953},{"x":324.4288823604584,"y":323.80863277950596},{"x":330.1851190328598,"y":314.6357821110756},{"x":334.1619029045105,"y":304.2468012455971},{"x":336.12103617191315,"y":294.20779072084736},{"x":338.0927973985672,"y":283.9055248681576},{"x":338.74841034412384,"y":272.77573021092724},{"x":260.76146980468184,"y":263.1945459429295},{"x":265.89059007167816,"y":260.37073077597927},{"x":272.02435222268105,"y":260.124662238744},{"x":277.91902008652687,"y":261.1794025782616},{"x":283.2442576289177,"y":263.57395942845653},{"x":305.61311280727386,"y":263.53190775552105},{"x":311.03955817222595,"y":261.3390071574719},{"x":317.3484830856323,"y":260.11370300093006},{"x":324.39085733890533,"y":260.55342193284343},{"x":329.6131602525711,"y":263.34477781095813},{"x":294.1689177751541,"y":276.0578458491833},{"x":294.0910659432411,"y":284.60743563213657},{"x":293.79795438051224,"y":292.4886583570511},{"x":293.3342087864876,"y":299.0591311339409},{"x":287.4225970506668,"y":300.58938663044285},{"x":290.1897984147072,"y":302.35011737385105},{"x":294.4143977165222,"y":303.4319332722695},{"x":298.62971901893616,"y":302.21711413898777},{"x":301.9501842260361,"y":301.3114898685486},{"x":269.1988005787134,"y":272.1574137095482},{"x":272.7045285999775,"y":270.833574045089},{"x":278.24089962244034,"y":270.82111893930744},{"x":283.33060455322266,"y":273.9352161768944},{"x":278.9782513678074,"y":275.6633411828549},{"x":273.0677897334099,"y":274.9888163093598},{"x":306.43324649333954,"y":274.1463497463734},{"x":311.48669731616974,"y":270.9408232454331},{"x":317.2671401500702,"y":271.0014146153004},{"x":321.12052023410797,"y":273.04276879468273},{"x":317.2139995098114,"y":275.7725465182335},{"x":311.281822681427,"y":275.4820482615502},{"x":277.7817610502243,"y":310.25390015878986},{"x":283.16760993003845,"y":309.13203405180286},{"x":290.7126435637474,"y":308.8409360531838},{"x":294.7775342464447,"y":309.4422331337006},{"x":298.55551266670227,"y":308.73483054676365},{"x":307.3698855638504,"y":310.28692780771564},{"x":314.2446427345276,"y":311.4532375816376},{"x":306.7817746400833,"y":318.6829243783028},{"x":300.91874945163727,"y":322.1013697270424},{"x":294.64789366722107,"y":322.99555008449863},{"x":289.2275020480156,"y":321.95475374498676},{"x":283.4413164258003,"y":318.63401751556705},{"x":278.51254960894585,"y":310.6321838978798},{"x":289.0830457806587,"y":311.44657765665363},{"x":294.83759093284607,"y":311.940667379287},{"x":300.5313273668289,"y":311.98795633354496},{"x":312.7565504312515,"y":311.62313889303516},{"x":300.40817058086395,"y":317.09120928087543},{"x":294.90167462825775,"y":317.9201293472321},{"x":289.45734149217606,"y":317.0948914174111}],[{"x":460.38254826515913,"y":226.86179668384213},{"x":461.868938408792,"y":239.01152131634373},{"x":464.2647043634206,"y":251.17557010727543},{"x":467.1906138891354,"y":260.80727634506843},{"x":471.15491477400064,"y":271.47034541445396},{"x":476.7369782626629,"y":279.180373466784},{"x":483.58512631058693,"y":284.52350709515235},{"x":492.1849847435951,"y":289.930207348639},{"x":505.92298462986946,"y":292.64942375498435},{"x":519.4783981442451,"y":289.30719820337913},{"x":529.468401491642,"y":284.01056245642326},{"x":537.524211525917,"y":278.7225001223626},{"x":544.3073756098747,"y":269.8340806730332},{"x":549.2690418362617,"y":259.2791630991044},{"x":552.0619995594025,"y":248.67283628063817},{"x":554.902642250061,"y":236.79101518469471},{"x":556.1653257608414,"y":224.22224742847104},{"x":465.896798262489,"y":214.67206279116292},{"x":471.0363594889641,"y":210.2688578732552},{"x":477.5265970826149,"y":209.1275984116139},{"x":484.0630382746458,"y":210.16732383447308},{"x":490.25007528066635,"y":212.04160619454998},{"x":513.647204041481,"y":211.24619571047444},{"x":520.0041263699532,"y":208.71019501166958},{"x":527.8550698161125,"y":207.4011332459988},{"x":535.9308312535286,"y":208.35862652140278},{"x":542.1029657125473,"y":212.3012642629685},{"x":501.67312091588974,"y":226.1231011219563},{"x":501.1025060117245,"y":234.8415866263451},{"x":500.3694550693035,"y":242.82483950930256},{"x":499.4915482401848,"y":250.10161349850316},{"x":494.7648799419403,"y":254.96258244591374},{"x":497.4780028760433,"y":256.4378731616082},{"x":501.41095703840256,"y":257.54796615915916},{"x":505.93551459908485,"y":256.1370541222634},{"x":509.78952783346176,"y":255.31036785917897},{"x":474.8673604428768,"y":224.3134863443913},{"x":478.4822910428047,"y":222.59203014450688},{"x":484.7031727582216,"y":222.44134343343396},{"x":490.5410310328007,"y":225.68867582755703},{"x":485.6805647313595,"y":227.69031397181172},{"x":479.1615803539753,"y":227.4026244350018},{"x":515.9134415388107,"y":224.7869722016396},{"x":521.3856536149979,"y":220.8381074138226},{"x":527.5773840546608,"y":220.89433822708745},{"x":532.6406374573708,"y":223.42468578296322},{"x":527.8728144168854,"y":226.2335638471188},{"x":521.3024847507477,"y":226.10652429538388},{"x":487.6637692153454,"y":269.516349054152},{"x":492.53167286515236,"y":268.0014941461625},{"x":498.14372247457504,"y":266.67949417906425},{"x":501.7443406879902,"y":267.2075164683404},{"x":505.5012784898281,"y":266.1977042921128},{"x":513.518238723278,"y":268.66181871967933},{"x":521.1821981072426,"y":269.4167241223397},{"x":514.5015707612038,"y":274.0143846638741},{"x":508.83223447203636,"y":276.6928109891953},{"x":502.9551590383053,"y":277.3477010496201},{"x":498.2226978838444,"y":276.7360996969285},{"x":492.5024300813675,"y":274.6970875628533},{"x":488.61172875761986,"y":269.74010888415},{"x":497.2433471083641,"y":270.23124519663475},{"x":502.03609746694565,"y":270.4881977208199},{"x":507.3701456785202,"y":270.1044398911538},{"x":519.4412326216698,"y":269.5110785134377},{"x":508.12173959612846,"y":271.158454335505},{"x":503.025142788887,"y":271.7769581444802},{"x":498.2751730978489,"y":271.27668622809074}],[{"x":575.4289033189416,"y":96.69157714613023},{"x":575.0089431256056,"y":108.45554218657556},{"x":575.3803079351783,"y":119.523285935002},{"x":576.5040263161063,"y":129.0130630827719},{"x":578.7652788180858,"y":139.67856867082656},{"x":584.0118243247271,"y":149.07947898872436},{"x":590.9061988592148,"y":157.3375187612349},{"x":600.3024386763573,"y":166.4939910865599},{"x":612.8085069060326,"y":171.9522984123999},{"x":625.6007747650146,"y":169.0268743134314},{"x":633.3883430957794,"y":162.2059975362593},{"x":640.3246023654938,"y":156.89696020849289},{"x":647.0983740091324,"y":148.92544485099853},{"x":652.5749768018723,"y":139.6908329821402},{"x":656.2116791009903,"y":130.98511714704574},{"x":659.2619792222977,"y":121.89113349683824},{"x":661.1847828626633,"y":111.414895633536},{"x":590.0366688966751,"y":97.08435080416741},{"x":597.0658887326717,"y":96.30971253879609},{"x":604.1793435811996,"y":96.69334376819673},{"x":610.7020010948181,"y":98.28364802368226},{"x":616.2523672580719,"y":100.93162317045274},{"x":635.8122313022614,"y":103.33054924972596},{"x":641.2868106365204,"y":101.84838841326776},{"x":647.3791997432709,"y":101.46065313585343},{"x":653.8026149272919,"y":102.04919339187684},{"x":657.612220287323,"y":103.94504354365411},{"x":624.058953166008,"y":112.44106070168557},{"x":623.6372038125992,"y":120.77518958814683},{"x":622.8375760316849,"y":128.73957468040527},{"x":621.6722447872162,"y":135.39052409895004},{"x":612.4750370383263,"y":134.6447736120993},{"x":615.8726144433022,"y":137.04947150714935},{"x":620.1995922327042,"y":138.6095454788977},{"x":624.4806340932846,"y":137.6083363390738},{"x":627.6450690031052,"y":136.39595527418197},{"x":597.6459639370441,"y":105.5917936063582},{"x":602.4066888689995,"y":105.266802111941},{"x":607.8324901461601,"y":105.51706613071504},{"x":612.0137300491333,"y":108.55084411271157},{"x":607.8006891012192,"y":109.66328666575494},{"x":601.8653501868248,"y":108.63868732221665},{"x":634.0711723566055,"y":111.0689470268065},{"x":639.5331710577011,"y":108.75894517667832},{"x":645.2980335950851,"y":109.52156786091867},{"x":648.0892332792282,"y":111.30079225667062},{"x":644.3416757583618,"y":113.4692985213572},{"x":638.816011428833,"y":112.59843675263467},{"x":598.5345310270786,"y":140.68911475904525},{"x":606.1386110186577,"y":141.5585543013388},{"x":614.658336520195,"y":142.71628488071502},{"x":618.7739044427872,"y":143.67765254743637},{"x":622.4421609640121,"y":143.17031312950195},{"x":630.0008096694946,"y":144.08841449983657},{"x":634.7515724897385,"y":144.11787945993484},{"x":627.894369840622,"y":151.4219178057486},{"x":622.2482999563217,"y":154.68191588886322},{"x":616.2834253907204,"y":155.18153191574157},{"x":610.9317002892494,"y":153.5865209198767},{"x":604.8389323949814,"y":149.88512880571426},{"x":599.5437785983086,"y":141.30464542873443},{"x":612.1623980402946,"y":144.4239004946524},{"x":618.0768132209778,"y":145.4545554614836},{"x":623.376920580864,"y":145.64860327013076},{"x":633.1576541662216,"y":144.4289133644873},{"x":622.6006407737732,"y":149.94578708179534},{"x":617.3486320972443,"y":150.5229176379019},{"x":611.8957805037498,"y":149.23444427021087}]]
\ No newline at end of file

test/tests/e2e/faceLandmarkNet.test.ts → test/tests/e2e/faceLandmark68Net.test.ts

 import * as tf from '@tensorflow/tfjs-core';
 
-import {
-  bufferToImage,
-  createFaceLandmarkNet,
-  Dimensions,
-  isTensor3D,
-  NetInput,
-  Point,
-  TMediaElement,
-  toNetInput,
-} from '../../../src';
+import { bufferToImage, Dimensions, isTensor3D, NetInput, Point, TMediaElement, toNetInput } from '../../../src';
 import { FaceLandmarks68 } from '../../../src/classes/FaceLandmarks68';
+import { createFaceLandmarkNet } from '../../../src/faceLandmarkNet';
 import { FaceLandmark68Net } from '../../../src/faceLandmarkNet/FaceLandmark68Net';
-import { describeWithNets, expectAllTensorsReleased, expectMaxDelta } from '../../utils';
+import { describeWithNets, expectAllTensorsReleased, expectMaxDelta, expectPointClose } from '../../utils';
 
 function getInputDims (input: tf.Tensor | TMediaElement): Dimensions {
   if (input instanceof tf.Tensor) {
@@ -22,7 +14,7 @@ function getInputDims (input: tf.Tensor | TMediaElement): Dimensions {
   return input
 }
 
-describe('faceLandmarkNet', () => {
+describe('faceLandmark68Net', () => {
 
   let imgEl1: HTMLImageElement
   let imgEl2: HTMLImageElement
@@ -43,71 +35,71 @@ describe('faceLandmarkNet', () => {
     faceLandmarkPositionsRect = await (await fetch('base/test/data/faceLandmarkPositionsRect.json')).json()
   })
 
-  describeWithNets('uncompressed weights', { withFaceLandmarkNet: { quantized: false } }, ({ faceLandmarkNet }) => {
+  describeWithNets('uncompressed weights', { withFaceLandmark68Net: { quantized: false } }, ({ faceLandmark68Net }) => {
 
     it('computes face landmarks for squared input', async () => {
       const { width, height } = imgEl1
 
-      const result = await faceLandmarkNet.detectLandmarks(imgEl1) as FaceLandmarks68
+      const result = await faceLandmark68Net.detectLandmarks(imgEl1) as FaceLandmarks68
       expect(result.getImageWidth()).toEqual(width)
       expect(result.getImageHeight()).toEqual(height)
       expect(result.getShift().x).toEqual(0)
       expect(result.getShift().y).toEqual(0)
-      result.getPositions().forEach(({ x, y }, i) => {
-        expectMaxDelta(x, faceLandmarkPositions1[i].x, 1)
-        expectMaxDelta(y, faceLandmarkPositions1[i].y, 1)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositions1[i]
+        expectPointClose(pt, { x, y }, 1)
       })
     })
 
     it('computes face landmarks for rectangular input', async () => {
       const { width, height } = imgElRect
 
-      const result = await faceLandmarkNet.detectLandmarks(imgElRect) as FaceLandmarks68
+      const result = await faceLandmark68Net.detectLandmarks(imgElRect) as FaceLandmarks68
       expect(result.getImageWidth()).toEqual(width)
       expect(result.getImageHeight()).toEqual(height)
       expect(result.getShift().x).toEqual(0)
       expect(result.getShift().y).toEqual(0)
-      result.getPositions().forEach(({ x, y }, i) => {
-        expectMaxDelta(x, faceLandmarkPositionsRect[i].x, 2)
-        expectMaxDelta(y, faceLandmarkPositionsRect[i].y, 2)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositionsRect[i]
+        expectPointClose(pt, { x, y }, 2)
       })
     })
 
   })
 
-  describeWithNets('quantized weights', { withFaceLandmarkNet: { quantized: true } }, ({ faceLandmarkNet }) => {
+  describeWithNets('quantized weights', { withFaceLandmark68Net: { quantized: true } }, ({ faceLandmark68Net }) => {
 
     it('computes face landmarks for squared input', async () => {
       const { width, height } = imgEl1
 
-      const result = await faceLandmarkNet.detectLandmarks(imgEl1) as FaceLandmarks68
+      const result = await faceLandmark68Net.detectLandmarks(imgEl1) as FaceLandmarks68
       expect(result.getImageWidth()).toEqual(width)
       expect(result.getImageHeight()).toEqual(height)
       expect(result.getShift().x).toEqual(0)
       expect(result.getShift().y).toEqual(0)
-      result.getPositions().forEach(({ x, y }, i) => {
-        expectMaxDelta(x, faceLandmarkPositions1[i].x, 2)
-        expectMaxDelta(y, faceLandmarkPositions1[i].y, 2)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositions1[i]
+        expectPointClose(pt, { x, y }, 2)
       })
     })
 
     it('computes face landmarks for rectangular input', async () => {
       const { width, height } = imgElRect
 
-      const result = await faceLandmarkNet.detectLandmarks(imgElRect) as FaceLandmarks68
+      const result = await faceLandmark68Net.detectLandmarks(imgElRect) as FaceLandmarks68
       expect(result.getImageWidth()).toEqual(width)
       expect(result.getImageHeight()).toEqual(height)
       expect(result.getShift().x).toEqual(0)
       expect(result.getShift().y).toEqual(0)
-      result.getPositions().forEach(({ x, y }, i) => {
-        expectMaxDelta(x, faceLandmarkPositionsRect[i].x, 6)
-        expectMaxDelta(y, faceLandmarkPositionsRect[i].y, 6)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositionsRect[i]
+        expectPointClose(pt, { x, y }, 6)
       })
     })
 
   })
 
-  describeWithNets('batch inputs', { withFaceLandmarkNet: { quantized: false } }, ({ faceLandmarkNet }) => {
+  describeWithNets('batch inputs', { withFaceLandmark68Net: { quantized: false } }, ({ faceLandmark68Net }) => {
 
     it('computes face landmarks for batch of image elements', async () => {
       const inputs = [imgEl1, imgEl2, imgElRect]
@@ -118,7 +110,7 @@ describe('faceLandmarkNet', () => {
         faceLandmarkPositionsRect
       ]
 
-      const results = await faceLandmarkNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      const results = await faceLandmark68Net.detectLandmarks(inputs) as FaceLandmarks68[]
       expect(Array.isArray(results)).toBe(true)
       expect(results.length).toEqual(3)
       results.forEach((result, batchIdx) => {
@@ -143,7 +135,7 @@ describe('faceLandmarkNet', () => {
         faceLandmarkPositionsRect
       ]
 
-      const results = await faceLandmarkNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      const results = await faceLandmark68Net.detectLandmarks(inputs) as FaceLandmarks68[]
       expect(Array.isArray(results)).toBe(true)
       expect(results.length).toEqual(3)
       results.forEach((result, batchIdx) => {
@@ -153,8 +145,8 @@ describe('faceLandmarkNet', () => {
         expect(result.getShift().x).toEqual(0)
         expect(result.getShift().y).toEqual(0)
         result.getPositions().forEach(({ x, y }, i) => {
-          expectMaxDelta(x, faceLandmarkPositions[batchIdx][i].x, 1)
-          expectMaxDelta(y, faceLandmarkPositions[batchIdx][i].y, 1)
+          expectMaxDelta(x, faceLandmarkPositions[batchIdx][i].x, 3)
+          expectMaxDelta(y, faceLandmarkPositions[batchIdx][i].y, 3)
         })
       })
     })
@@ -168,7 +160,7 @@ describe('faceLandmarkNet', () => {
         faceLandmarkPositionsRect
       ]
 
-      const results = await faceLandmarkNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      const results = await faceLandmark68Net.detectLandmarks(inputs) as FaceLandmarks68[]
       expect(Array.isArray(results)).toBe(true)
       expect(results.length).toEqual(3)
       results.forEach((result, batchIdx) => {
@@ -178,15 +170,15 @@ describe('faceLandmarkNet', () => {
         expect(result.getShift().x).toEqual(0)
         expect(result.getShift().y).toEqual(0)
         result.getPositions().forEach(({ x, y }, i) => {
-          expectMaxDelta(x, faceLandmarkPositions[batchIdx][i].x, 1)
-          expectMaxDelta(y, faceLandmarkPositions[batchIdx][i].y, 1)
+          expectMaxDelta(x, faceLandmarkPositions[batchIdx][i].x, 3)
+          expectMaxDelta(y, faceLandmarkPositions[batchIdx][i].y, 3)
         })
       })
     })
 
   })
 
-  describeWithNets('no memory leaks', { withFaceLandmarkNet: { quantized: true } }, ({ faceLandmarkNet }) => {
+  describeWithNets('no memory leaks', { withFaceLandmark68Net: { quantized: true } }, ({ faceLandmark68Net }) => {
 
     describe('NeuralNetwork, uncompressed model', () => {
 
@@ -218,7 +210,7 @@ describe('faceLandmarkNet', () => {
       it('single image element', async () => {
         await expectAllTensorsReleased(async () => {
           const netInput = new NetInput([imgEl1])
-          const outTensor = await faceLandmarkNet.forwardInput(netInput)
+          const outTensor = await faceLandmark68Net.forwardInput(netInput)
           outTensor.dispose()
         })
       })
@@ -226,7 +218,7 @@ describe('faceLandmarkNet', () => {
       it('multiple image elements', async () => {
         await expectAllTensorsReleased(async () => {
           const netInput = new NetInput([imgEl1, imgEl1, imgEl1])
-          const outTensor = await faceLandmarkNet.forwardInput(netInput)
+          const outTensor = await faceLandmark68Net.forwardInput(netInput)
           outTensor.dispose()
         })
       })
@@ -236,7 +228,7 @@ describe('faceLandmarkNet', () => {
 
         await expectAllTensorsReleased(async () => {
           const netInput = new NetInput([tensor])
-          const outTensor = await faceLandmarkNet.forwardInput(netInput)
+          const outTensor = await faceLandmark68Net.forwardInput(netInput)
           outTensor.dispose()
         })
 
@@ -248,7 +240,7 @@ describe('faceLandmarkNet', () => {
 
         await expectAllTensorsReleased(async () => {
           const netInput = new NetInput(tensors)
-          const outTensor = await faceLandmarkNet.forwardInput(netInput)
+          const outTensor = await faceLandmark68Net.forwardInput(netInput)
           outTensor.dispose()
         })
 
@@ -259,7 +251,7 @@ describe('faceLandmarkNet', () => {
         const tensor = tf.tidy(() => tf.fromPixels(imgEl1).expandDims()) as tf.Tensor4D
 
         await expectAllTensorsReleased(async () => {
-          const outTensor = await faceLandmarkNet.forwardInput(await toNetInput(tensor))
+          const outTensor = await faceLandmark68Net.forwardInput(await toNetInput(tensor))
           outTensor.dispose()
         })
 
@@ -271,7 +263,7 @@ describe('faceLandmarkNet', () => {
           .map(el => tf.tidy(() => tf.fromPixels(el).expandDims())) as tf.Tensor4D[]
 
         await expectAllTensorsReleased(async () => {
-          const outTensor = await faceLandmarkNet.forwardInput(await toNetInput(tensors))
+          const outTensor = await faceLandmark68Net.forwardInput(await toNetInput(tensors))
           outTensor.dispose()
         })
 
@@ -284,13 +276,13 @@ describe('faceLandmarkNet', () => {
 
       it('single image element', async () => {
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks(imgEl1)
+          await faceLandmark68Net.detectLandmarks(imgEl1)
         })
       })
 
       it('multiple image elements', async () => {
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks([imgEl1, imgEl1, imgEl1])
+          await faceLandmark68Net.detectLandmarks([imgEl1, imgEl1, imgEl1])
         })
       })
 
@@ -298,7 +290,7 @@ describe('faceLandmarkNet', () => {
         const tensor = tf.fromPixels(imgEl1)
 
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks(tensor)
+          await faceLandmark68Net.detectLandmarks(tensor)
         })
 
         tensor.dispose()
@@ -309,7 +301,7 @@ describe('faceLandmarkNet', () => {
 
 
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks(tensors)
+          await faceLandmark68Net.detectLandmarks(tensors)
         })
 
         tensors.forEach(t => t.dispose())
@@ -319,7 +311,7 @@ describe('faceLandmarkNet', () => {
         const tensor = tf.tidy(() => tf.fromPixels(imgEl1).expandDims()) as tf.Tensor4D
 
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks(tensor)
+          await faceLandmark68Net.detectLandmarks(tensor)
         })
 
         tensor.dispose()
@@ -330,7 +322,7 @@ describe('faceLandmarkNet', () => {
           .map(el => tf.tidy(() => tf.fromPixels(el).expandDims())) as tf.Tensor4D[]
 
         await expectAllTensorsReleased(async () => {
-          await faceLandmarkNet.detectLandmarks(tensors)
+          await faceLandmark68Net.detectLandmarks(tensors)
         })
 
         tensors.forEach(t => t.dispose())

test/tests/e2e/faceLandmark68TinyNet.test.ts

+import * as tf from '@tensorflow/tfjs-core';
+
+import { bufferToImage, Dimensions, isTensor3D, NetInput, Point, TMediaElement, toNetInput } from '../../../src';
+import { FaceLandmarks68 } from '../../../src/classes/FaceLandmarks68';
+import { createFaceLandmarkNet } from '../../../src/faceLandmarkNet';
+import { FaceLandmark68TinyNet } from '../../../src/faceLandmarkNet/FaceLandmark68TinyNet';
+import { describeWithNets, expectAllTensorsReleased, expectMaxDelta, expectPointClose } from '../../utils';
+
+function getInputDims (input: tf.Tensor | TMediaElement): Dimensions {
+  if (input instanceof tf.Tensor) {
+    const [height, width] = input.shape.slice(isTensor3D(input) ? 0 : 1)
+    return { width, height }
+  }
+  return input
+}
+
+describe('faceLandmark68TinyNet', () => {
+
+  let imgEl1: HTMLImageElement
+  let imgEl2: HTMLImageElement
+  let imgElRect: HTMLImageElement
+  let faceLandmarkPositions1: Point[]
+  let faceLandmarkPositions2: Point[]
+  let faceLandmarkPositionsRect: Point[]
+
+  beforeAll(async () => {
+    const img1 = await (await fetch('base/test/images/face1.png')).blob()
+    imgEl1 = await bufferToImage(img1)
+    const img2 = await (await fetch('base/test/images/face2.png')).blob()
+    imgEl2 = await bufferToImage(img2)
+    const imgRect = await (await fetch('base/test/images/face_rectangular.png')).blob()
+    imgElRect = await bufferToImage(imgRect)
+    faceLandmarkPositions1 = await (await fetch('base/test/data/faceLandmarkPositions1Tiny.json')).json()
+    faceLandmarkPositions2 = await (await fetch('base/test/data/faceLandmarkPositions2Tiny.json')).json()
+    faceLandmarkPositionsRect = await (await fetch('base/test/data/faceLandmarkPositionsRectTiny.json')).json()
+  })
+
+  describeWithNets('uncompressed weights', { withFaceLandmark68TinyNet: { quantized: false } }, ({ faceLandmark68TinyNet }) => {
+
+    it('computes face landmarks for squared input', async () => {
+      const { width, height } = imgEl1
+
+      const result = await faceLandmark68TinyNet.detectLandmarks(imgEl1) as FaceLandmarks68
+      expect(result.getImageWidth()).toEqual(width)
+      expect(result.getImageHeight()).toEqual(height)
+      expect(result.getShift().x).toEqual(0)
+      expect(result.getShift().y).toEqual(0)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositions1[i]
+        expectPointClose(pt, { x, y }, 5)
+      })
+    })
+
+    it('computes face landmarks for rectangular input', async () => {
+      const { width, height } = imgElRect
+
+      const result = await faceLandmark68TinyNet.detectLandmarks(imgElRect) as FaceLandmarks68
+      expect(result.getImageWidth()).toEqual(width)
+      expect(result.getImageHeight()).toEqual(height)
+      expect(result.getShift().x).toEqual(0)
+      expect(result.getShift().y).toEqual(0)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositionsRect[i]
+        expectPointClose(pt, { x, y }, 5)
+      })
+    })
+
+  })
+
+  describeWithNets('quantized weights', { withFaceLandmark68TinyNet: { quantized: true } }, ({ faceLandmark68TinyNet }) => {
+
+    it('computes face landmarks for squared input', async () => {
+      const { width, height } = imgEl1
+
+      const result = await faceLandmark68TinyNet.detectLandmarks(imgEl1) as FaceLandmarks68
+      expect(result.getImageWidth()).toEqual(width)
+      expect(result.getImageHeight()).toEqual(height)
+      expect(result.getShift().x).toEqual(0)
+      expect(result.getShift().y).toEqual(0)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositions1[i]
+        expectPointClose(pt, { x, y }, 5)
+      })
+    })
+
+    it('computes face landmarks for rectangular input', async () => {
+      const { width, height } = imgElRect
+
+      const result = await faceLandmark68TinyNet.detectLandmarks(imgElRect) as FaceLandmarks68
+      expect(result.getImageWidth()).toEqual(width)
+      expect(result.getImageHeight()).toEqual(height)
+      expect(result.getShift().x).toEqual(0)
+      expect(result.getShift().y).toEqual(0)
+      result.getPositions().forEach((pt, i) => {
+        const { x, y } = faceLandmarkPositionsRect[i]
+        expectPointClose(pt, { x, y }, 5)
+      })
+    })
+
+  })
+
+  describeWithNets('batch inputs', { withFaceLandmark68TinyNet: { quantized: false } }, ({ faceLandmark68TinyNet }) => {
+
+    it('computes face landmarks for batch of image elements', async () => {
+      const inputs = [imgEl1, imgEl2, imgElRect]
+
+      const faceLandmarkPositions = [
+        faceLandmarkPositions1,
+        faceLandmarkPositions2,
+        faceLandmarkPositionsRect
+      ]
+
+      const results = await faceLandmark68TinyNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      expect(Array.isArray(results)).toBe(true)
+      expect(results.length).toEqual(3)
+      results.forEach((result, batchIdx) => {
+        const { width, height } = getInputDims(inputs[batchIdx])
+        expect(result.getImageWidth()).toEqual(width)
+        expect(result.getImageHeight()).toEqual(height)
+        expect(result.getShift().x).toEqual(0)
+        expect(result.getShift().y).toEqual(0)
+        result.getPositions().forEach((pt, i) => {
+          const { x, y } = faceLandmarkPositions[batchIdx][i]
+          expectPointClose(pt, { x, y }, 5)
+        })
+      })
+    })
+
+    it('computes face landmarks for batch of tf.Tensor3D', async () => {
+      const inputs = [imgEl1, imgEl2, imgElRect].map(el => tf.fromPixels(el))
+
+      const faceLandmarkPositions = [
+        faceLandmarkPositions1,
+        faceLandmarkPositions2,
+        faceLandmarkPositionsRect
+      ]
+
+      const results = await faceLandmark68TinyNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      expect(Array.isArray(results)).toBe(true)
+      expect(results.length).toEqual(3)
+      results.forEach((result, batchIdx) => {
+        const { width, height } = getInputDims(inputs[batchIdx])
+        expect(result.getImageWidth()).toEqual(width)
+        expect(result.getImageHeight()).toEqual(height)
+        expect(result.getShift().x).toEqual(0)
+        expect(result.getShift().y).toEqual(0)
+        result.getPositions().forEach((pt, i) => {
+          const { x, y } = faceLandmarkPositions[batchIdx][i]
+          expectPointClose(pt, { x, y }, 3)
+        })
+      })
+    })
+
+    it('computes face landmarks for batch of mixed inputs', async () => {
+      const inputs = [imgEl1, tf.fromPixels(imgEl2), tf.fromPixels(imgElRect)]
+
+      const faceLandmarkPositions = [
+        faceLandmarkPositions1,
+        faceLandmarkPositions2,
+        faceLandmarkPositionsRect
+      ]
+
+      const results = await faceLandmark68TinyNet.detectLandmarks(inputs) as FaceLandmarks68[]
+      expect(Array.isArray(results)).toBe(true)
+      expect(results.length).toEqual(3)
+      results.forEach((result, batchIdx) => {
+        const { width, height } = getInputDims(inputs[batchIdx])
+        expect(result.getImageWidth()).toEqual(width)
+        expect(result.getImageHeight()).toEqual(height)
+        expect(result.getShift().x).toEqual(0)
+        expect(result.getShift().y).toEqual(0)
+        result.getPositions().forEach((pt, i) => {
+          const { x, y } = faceLandmarkPositions[batchIdx][i]
+          expectPointClose(pt, { x, y }, 3)
+        })
+      })
+    })
+
+  })
+
+  describeWithNets('no memory leaks', { withFaceLandmark68TinyNet: { quantized: true } }, ({ faceLandmark68TinyNet }) => {
+
+    describe('NeuralNetwork, uncompressed model', () => {
+
+      it('disposes all param tensors', async () => {
+        await expectAllTensorsReleased(async () => {
+          const res = await fetch('base/weights_uncompressed/face_landmark_68_model.weights')
+          const weights = new Float32Array(await res.arrayBuffer())
+          const net = createFaceLandmarkNet(weights)
+          net.dispose()
+        })
+      })
+
+    })
+
+    describe('NeuralNetwork, quantized model', () => {
+
+      it('disposes all param tensors', async () => {
+        await expectAllTensorsReleased(async () => {
+          const net = new FaceLandmark68TinyNet()
+          await net.load('base/weights')
+          net.dispose()
+        })
+      })
+
+    })
+
+    describe('forwardInput', () => {
+
+      it('single image element', async () => {
+        await expectAllTensorsReleased(async () => {
+          const netInput = new NetInput([imgEl1])
+          const outTensor = await faceLandmark68TinyNet.forwardInput(netInput)
+          outTensor.dispose()
+        })
+      })
+
+      it('multiple image elements', async () => {
+        await expectAllTensorsReleased(async () => {
+          const netInput = new NetInput([imgEl1, imgEl1, imgEl1])
+          const outTensor = await faceLandmark68TinyNet.forwardInput(netInput)
+          outTensor.dispose()
+        })
+      })
+
+      it('single tf.Tensor3D', async () => {
+        const tensor = tf.fromPixels(imgEl1)
+
+        await expectAllTensorsReleased(async () => {
+          const netInput = new NetInput([tensor])
+          const outTensor = await faceLandmark68TinyNet.forwardInput(netInput)
+          outTensor.dispose()
+        })
+
+        tensor.dispose()
+      })
+
+      it('multiple tf.Tensor3Ds', async () => {
+        const tensors = [imgEl1, imgEl1, imgEl1].map(el => tf.fromPixels(el))
+
+        await expectAllTensorsReleased(async () => {
+          const netInput = new NetInput(tensors)
+          const outTensor = await faceLandmark68TinyNet.forwardInput(netInput)
+          outTensor.dispose()
+        })
+
+        tensors.forEach(t => t.dispose())
+      })
+
+      it('single batch size 1 tf.Tensor4Ds', async () => {
+        const tensor = tf.tidy(() => tf.fromPixels(imgEl1).expandDims()) as tf.Tensor4D
+
+        await expectAllTensorsReleased(async () => {
+          const outTensor = await faceLandmark68TinyNet.forwardInput(await toNetInput(tensor))
+          outTensor.dispose()
+        })
+
+        tensor.dispose()
+      })
+
+      it('multiple batch size 1 tf.Tensor4Ds', async () => {
+        const tensors = [imgEl1, imgEl1, imgEl1]
+          .map(el => tf.tidy(() => tf.fromPixels(el).expandDims())) as tf.Tensor4D[]
+
+        await expectAllTensorsReleased(async () => {
+          const outTensor = await faceLandmark68TinyNet.forwardInput(await toNetInput(tensors))
+          outTensor.dispose()
+        })
+
+        tensors.forEach(t => t.dispose())
+      })
+
+    })
+
+    describe('detectLandmarks', () => {
+
+      it('single image element', async () => {
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks(imgEl1)
+        })
+      })
+
+      it('multiple image elements', async () => {
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks([imgEl1, imgEl1, imgEl1])
+        })
+      })
+
+      it('single tf.Tensor3D', async () => {
+        const tensor = tf.fromPixels(imgEl1)
+
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks(tensor)
+        })
+
+        tensor.dispose()
+      })
+
+      it('multiple tf.Tensor3Ds', async () => {
+        const tensors = [imgEl1, imgEl1, imgEl1].map(el => tf.fromPixels(el))
+
+
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks(tensors)
+        })
+
+        tensors.forEach(t => t.dispose())
+      })
+
+      it('single batch size 1 tf.Tensor4Ds', async () => {
+        const tensor = tf.tidy(() => tf.fromPixels(imgEl1).expandDims()) as tf.Tensor4D
+
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks(tensor)
+        })
+
+        tensor.dispose()
+      })
+
+      it('multiple batch size 1 tf.Tensor4Ds', async () => {
+        const tensors = [imgEl1, imgEl1, imgEl1]
+          .map(el => tf.tidy(() => tf.fromPixels(el).expandDims())) as tf.Tensor4D[]
+
+        await expectAllTensorsReleased(async () => {
+          await faceLandmark68TinyNet.detectLandmarks(tensors)
+        })
+
+        tensors.forEach(t => t.dispose())
+      })
+
+    })
+  })
+
+})
+

test/utils.ts

 import * as tf from '@tensorflow/tfjs-core';
 
-import {
-  FaceDetectionNet,
-  FaceLandmarkNet,
-  FaceRecognitionNet,
-  IPoint,
-  IRect,
-  Mtcnn,
-  NeuralNetwork,
-  TinyYolov2,
-} from '../src/';
+import { FaceDetectionNet, FaceRecognitionNet, IPoint, IRect, Mtcnn, NeuralNetwork, TinyYolov2 } from '../src/';
 import { allFacesMtcnnFactory, allFacesSsdMobilenetv1Factory, allFacesTinyYolov2Factory } from '../src/allFacesFactory';
 import { FaceDetection } from '../src/classes/FaceDetection';
 import { FaceLandmarks } from '../src/classes/FaceLandmarks';
 import { FullFaceDescription } from '../src/classes/FullFaceDescription';
+import { FaceLandmark68Net } from '../src/faceLandmarkNet/FaceLandmark68Net';
+import { FaceLandmark68TinyNet } from '../src/faceLandmarkNet/FaceLandmark68TinyNet';
 import { allFacesMtcnnFunction, allFacesSsdMobilenetv1Function, allFacesTinyYolov2Function } from '../src/globalApi';
 
 jasmine.DEFAULT_TIMEOUT_INTERVAL = 60000
@@ -106,7 +99,8 @@ export type InjectNetArgs = {
   allFacesTinyYolov2: allFacesTinyYolov2Function
   allFacesMtcnn: allFacesMtcnnFunction
   faceDetectionNet: FaceDetectionNet
-  faceLandmarkNet: FaceLandmarkNet
+  faceLandmark68Net: FaceLandmark68Net
+  faceLandmark68TinyNet: FaceLandmark68TinyNet
   faceRecognitionNet: FaceRecognitionNet
   mtcnn: Mtcnn
   tinyYolov2: TinyYolov2
@@ -118,7 +112,8 @@ export type DescribeWithNetsOptions = {
   withAllFacesTinyYolov2?: boolean
   withAllFacesMtcnn?: boolean
   withFaceDetectionNet?: WithNetOptions
-  withFaceLandmarkNet?: WithNetOptions
+  withFaceLandmark68Net?: WithNetOptions
+  withFaceLandmark68TinyNet?: WithNetOptions
   withFaceRecognitionNet?: WithNetOptions
   withMtcnn?: WithNetOptions
   withTinyYolov2?: WithTinyYolov2Options
@@ -148,12 +143,13 @@ export function describeWithNets(
 
   describe(description, () => {
     let faceDetectionNet: FaceDetectionNet = new FaceDetectionNet()
-    let faceLandmarkNet: FaceLandmarkNet = new FaceLandmarkNet()
+    let faceLandmark68Net: FaceLandmark68Net = new FaceLandmark68Net()
+    let faceLandmark68TinyNet: FaceLandmark68TinyNet = new FaceLandmark68TinyNet()
     let faceRecognitionNet: FaceRecognitionNet = new FaceRecognitionNet()
     let mtcnn: Mtcnn = new Mtcnn()
     let tinyYolov2: TinyYolov2 = new TinyYolov2(options.withTinyYolov2 && options.withTinyYolov2.withSeparableConv)
-    let allFacesSsdMobilenetv1 = allFacesSsdMobilenetv1Factory(faceDetectionNet, faceLandmarkNet, faceRecognitionNet)
-    let allFacesTinyYolov2 = allFacesTinyYolov2Factory(tinyYolov2, faceLandmarkNet, faceRecognitionNet)
+    let allFacesSsdMobilenetv1 = allFacesSsdMobilenetv1Factory(faceDetectionNet, faceLandmark68Net, faceRecognitionNet)
+    let allFacesTinyYolov2 = allFacesTinyYolov2Factory(tinyYolov2, faceLandmark68Net, faceRecognitionNet)
     let allFacesMtcnn = allFacesMtcnnFactory(mtcnn, faceRecognitionNet)
 
     beforeAll(async () => {
@@ -162,7 +158,8 @@ export function describeWithNets(
         withAllFacesTinyYolov2,
         withAllFacesMtcnn,
         withFaceDetectionNet,
-        withFaceLandmarkNet,
+        withFaceLandmark68Net,
+        withFaceLandmark68TinyNet,
         withFaceRecognitionNet,
         withMtcnn,
         withTinyYolov2
@@ -175,10 +172,17 @@ export function describeWithNets(
         )
       }
 
-      if (withFaceLandmarkNet || withAllFacesSsdMobilenetv1 || withAllFacesTinyYolov2) {
-        await initNet<FaceLandmarkNet>(
-          faceLandmarkNet,
-          !!withFaceLandmarkNet && !withFaceLandmarkNet.quantized && 'face_landmark_68_model.weights'
+      if (withFaceLandmark68Net || withAllFacesSsdMobilenetv1 || withAllFacesTinyYolov2) {
+        await initNet<FaceLandmark68Net>(
+          faceLandmark68Net,
+          !!withFaceLandmark68Net && !withFaceLandmark68Net.quantized && 'face_landmark_68_model.weights'
+        )
+      }
+
+      if (withFaceLandmark68TinyNet) {
+        await initNet<FaceLandmark68TinyNet>(
+          faceLandmark68TinyNet,
+          !!withFaceLandmark68TinyNet && !withFaceLandmark68TinyNet.quantized && 'face_landmark_68_tiny_model.weights'
         )
       }
 
@@ -208,7 +212,7 @@ export function describeWithNets(
 
     afterAll(() => {
       faceDetectionNet && faceDetectionNet.dispose()
-      faceLandmarkNet && faceLandmarkNet.dispose()
+      faceLandmark68Net && faceLandmark68Net.dispose()
       faceRecognitionNet && faceRecognitionNet.dispose()
       mtcnn && mtcnn.dispose(),
       tinyYolov2 && tinyYolov2.dispose()
@@ -219,7 +223,8 @@ export function describeWithNets(
       allFacesTinyYolov2,
       allFacesMtcnn,
       faceDetectionNet,
-      faceLandmarkNet,
+      faceLandmark68Net,
+      faceLandmark68TinyNet,
       faceRecognitionNet,
       mtcnn,
       tinyYolov2

tools/quantize/views/quantizeModel.html

@@ -2,7 +2,6 @@
 <html>
 <head>
   <script src="face-api.js"></script>
-  <script src="commons.js"></script>
   <script src="FileSaver.js"></script>
   <script src="quantization.js"></script>
 </head>
@@ -10,14 +9,20 @@
   <script>
     tf = faceapi.tf
 
-    const modelName = 'tiny_yolov2_separable_conv'
-    const uncompressedWeightsUri = `tiny_yolov2_separable_conv_model_v1.weights`
-    const net = new faceapi.TinyYolov2(true)
+    const uncompressedWeightsUri = `face_landmark_68_model.weights`
+    const net = new faceapi.FaceLandmark68LargeNet()
 
-    async function loadNetWeights(uri) {
-      return new Float32Array(await (await fetch(uri)).arrayBuffer())
+    async function load() {
+      await net.load(new Float32Array(await (await fetch(uncompressedWeightsUri)).arrayBuffer()))
+      console.log('net loaded')
     }
 
+    function getNamedTensors() {
+      return net.getParamList().map(({ path, tensor }) => ({ name: path, tensor }))
+    }
+
+    const modelName = 'face_landmark_68'
+
     function makeShards(weightArray) {
       const maxLength = 4096 * 1024
 
@@ -41,25 +46,34 @@
 
     async function quantizeAndSave() {
 
-      await net.load(await loadNetWeights(uncompressedWeightsUri))
-
       const quantizedTensorArrays = []
       const weightEntries = []
 
-      net.getParamList().forEach(({ path, tensor }) => {
-        const { scale, min, qdata } = quantizeWeights(tensor)
+      getNamedTensors().forEach(({ name, tensor, isSkipQuantization }) => {
 
         const weightEntry = {
-          name : path,
+          name,
           shape: tensor.shape,
-          dtype: tensor.dtype,
-          quantization: { dtype: 'uint8', scale, min }
+          dtype: tensor.dtype
         }
 
-        console.log({ scale, min })
+        if (isSkipQuantization) {
+          quantizedTensorArrays.push(tensor.dataSync())
+          weightEntries.push(weightEntry)
+          return
+        }
+
+        const { scale, min, qdata } = quantizeWeights(tensor)
+
+        console.log(name, { scale, min })
+
+        const quantizedWeightEntry = {
+          ...weightEntry,
+          quantization: { dtype: 'uint8', scale, min }
+        }
 
         quantizedTensorArrays.push(qdata)
-        weightEntries.push(weightEntry)
+        weightEntries.push(quantizedWeightEntry)
       })
 
       const quantizedWeights = quantizedTensorArrays
@@ -83,6 +97,9 @@
 
       saveAs(new Blob([JSON.stringify(weightManifest)]), `${modelName}_model-weights_manifest.json`)
     }
+
+    load()
+
   </script>
 </body>
 </html>
\ No newline at end of file

tools/train/faceLandmarks/faceLandmarksTrain.js

-async function trainStep(batchCreators) {
-  await promiseSequential(batchCreators.map((batchCreator, dataIdx) => async () => {
-
-    const { batchInput, landmarksBatchTensor } = await batchCreator()
-
-    let ts = Date.now()
-    const cost = optimizer.minimize(() => {
-      const out = window.trainNet.forwardInput(batchInput.managed())
-      const loss = lossFunction(
-        landmarksBatchTensor,
-        out
-      )
-      return tf.sum(out)
-    }, true)
-
-    ts = Date.now() - ts
-    console.log(`loss[${dataIdx}]: ${await cost.data()}, ${ts} ms (${ts / batchInput.batchSize} ms / batch element)`)
-
-    landmarksBatchTensor.dispose()
-    cost.dispose()
-
-    await tf.nextFrame()
-    console.log(tf.memory())
-  }))
-}
-
-function createBatchCreators(data, batchSize) {
-  if (batchSize < 1) {
-    throw new Error('invalid batch size: ' + batchSize)
-  }
-
-  const batches = []
-  const pushToBatch = (remaining) => {
-    if (remaining.length) {
-      batches.push(remaining.slice(0, batchSize))
-      pushToBatch(remaining.slice(batchSize))
-    }
-    return batches
-  }
-
-  pushToBatch(data)
-
-  const batchCreators = batches.map(dataForBatch => async () => {
-    const imgs = dataForBatch.map(d => d.img)
-    const allLandmarks = dataForBatch.map(d => landmarkPositionsToArray(d.landmarks))
-
-    const batchInput = await faceapi.toNetInput(imgs)
-    const landmarksBatchTensor = tf.tidy(() => {
-      const landmarkTensors = allLandmarks.map(arr => tf.tensor2d(arr, [1, 136]))
-      return tf.stack(landmarkTensors, 0).as2D(-1, 136)
-    })
-
-    return {
-      batchInput,
-      landmarksBatchTensor
-    }
-  })
-
-  return batchCreators
-}
-
-function landmarkPositionsToArray(landmarks) {
-  return landmarks.getRelativePositions().map(pt => [pt.x, pt.y])
-    .reduce((flat, arr) => flat.concat(arr))
-}
-
-function toFaceLandmarks(landmarks, { naturalWidth, naturalHeight }) {
-  return new faceapi.FaceLandmarks68(
-    landmarks.map(l => new faceapi.Point(l.x / naturalWidth, l.y / naturalHeight)),
-    { width: naturalWidth, height: naturalHeight }
-  )
-}
-
-async function loadImagesInBatch(allLandmarks, offset = 0) {
-  return Promise.all(allLandmarks.map(async (landmarks, i) => {
-    const imgUri = `/train_images/${i + offset}.png`
-    const img = await faceapi.bufferToImage(await fetchImage(imgUri))
-    return {
-      imgUri,
-      img,
-      landmarks: toFaceLandmarks(landmarks, img)
-    }
-  }))
-}
-
-async function getTrainData() {
-  const landmarksJson = (await (await fetch('/train_landmarks.json')).json())
-
-  const numLandmarks = Object.keys(landmarksJson).length
-  const allLandmarks = Array.from(
-    new Array(numLandmarks),
-    (_, i) => landmarksJson[i]
-  )
-
-  return await loadImagesInBatch(allLandmarks.slice(0, 100))
-/**
-  const batch1 = await loadImagesInBatch(allLandmarks.slice(0, 4000))
-  const batch2 = await loadImagesInBatch(allLandmarks.slice(4000), 4000)
-
-  return batch1.concat(batch2)
-  */
-}
\ No newline at end of file

tools/train/faceLandmarks/faceLandmarksUi.js

-function onKeyDown(e) {
-  e.target.value = (
-    parseInt(e.target.value) + (e.keyCode === 38 ? 1 : (e.keyCode === 40 ? -1 : 0))
-  ) || e.target.value || 0
-
-  const uri = '/train_images/' + e.target.value + '.png'
-
-  console.log(uri)
-  onSelectionChanged(uri)
-}
-
-function onChangeDrawLines(e) {
-  window.drawLines = $(e.target).prop('checked')
-  redraw()
-}
-
-async function onSelectionChanged(uri) {
-  const imgBuf = await fetchImage(uri)
-  window.currentImg = await faceapi.bufferToImage(imgBuf)
-  window.landmarks = await Promise.all(window.nets.map(
-    async net => net.detectLandmarks(window.currentImg)
-  ))
-  redraw(uri)
-}
-
-function drawLandmarkCanvas(img, landmarks, drawOpts) {
-  const canvas = faceapi.createCanvasFromMedia(img)
-  $('#faceContainer').append(canvas)
-
-  faceapi.drawLandmarks(
-    canvas,
-    landmarks,
-    drawOpts
-  )
-}
-
-function redraw(uri) {
-  $('#faceContainer').empty()
-
-  const drawOpts = { lineWidth: window.drawLines ? 2 : 4, drawLines: window.drawLines }
-  window.landmarks.forEach(landmarks => {
-    drawLandmarkCanvas(window.currentImg, landmarks, drawOpts)
-  })
-
-  const trainImgAndLandmarks = window.trainData.find(t => t.imgUri === uri)
-  if (trainImgAndLandmarks) {
-    drawLandmarkCanvas(trainImgAndLandmarks.img, trainImgAndLandmarks.landmarks, drawOpts)
-  }
-}
-
-async function loadNet(file) {
-  const res = await fetch(file)
-  const weights = new Float32Array(await res.arrayBuffer())
-  return faceapi.faceLandmarkNet(weights)
-}
-
-async function init() {
-  //await faceapi.loadFaceLandmarkModel('/')
-  //window.nets.push(faceapi.landmarkNet)
-  //window.nets.push(await loadNet('retrained/landmarks_v0.weights'))
-  //window.nets.push(await loadNet('retrained/landmarks_v2.weights'))
-
-  window.trainNet = await loadNet('/tmp/retrained/landmarks_v9.weights')
-  window.nets.push(trainNet)
-
-  $('#loader').hide()
-  await run()
-  await onSelectionChanged($('#selectList select').val())
-}
-
-$(document).ready(function() {
-  $('#imgByNr').keydown(onKeyDown);
-  renderFaceImageSelectList(
-    '#selectList',
-    onSelectionChanged,
-    { className: 'sheldon', imageIdx: 1 }
-  )
-  init()
-})
\ No newline at end of file

tools/train/faceLandmarks/initWeightsDensenet.html

+<!DOCTYPE html>
+<html>
+<head>
+  <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@0.12.0"> </script>
+  <script src="../node_modules/file-saver/FileSaver.js"></script>
+</head>
+<body>
+
+  <div class="row side-by-side">
+    <button
+      class="waves-effect waves-light btn"
+      onclick="save()"
+    >
+      Save
+    </button>
+    <button
+      class="waves-effect waves-light btn"
+      onclick="save(true)"
+    >
+      Save Tiny
+    </button>
+  </div>
+
+  <script>
+    function toDataArray(tensor) {
+      return Array.from(tensor.dataSync())
+    }
+
+    function flatten(arrs) {
+      return arrs.reduce((flat, arr) => flat.concat(arr))
+    }
+
+    function initWeights(initializer, isTiny) {
+      function initConvWeights(inChannels, outChannels) {
+        return flatten(
+          [
+            // filter
+            initializer.apply([3, 3, inChannels, outChannels]),
+            // bias
+            tf.zeros([outChannels])
+          ]
+            .map(toDataArray)
+        )
+      }
+
+      function initSeparableConvWeights(inChannels, outChannels) {
+        return flatten(
+          [
+            // depthwise filter
+            initializer.apply([3, 3, inChannels, 1]),
+            // pointwise filter
+            initializer.apply([1, 1, inChannels, outChannels]),
+            // bias
+            tf.zeros([outChannels])
+          ]
+            .map(toDataArray)
+        )
+      }
+
+      const separableConvWeights = isTiny
+        ? flatten([
+          initConvWeights(3, 32),
+          initSeparableConvWeights(32, 32),
+          initSeparableConvWeights(32, 32),
+          initSeparableConvWeights(32, 64),
+          initSeparableConvWeights(64, 64),
+          initSeparableConvWeights(64, 64),
+          initSeparableConvWeights(64, 128),
+          initSeparableConvWeights(128, 128),
+          initSeparableConvWeights(128, 128)
+        ])
+        : flatten([
+          initConvWeights(3, 32),
+          initSeparableConvWeights(32, 32),
+          initSeparableConvWeights(32, 32),
+          initSeparableConvWeights(32, 32),
+          initSeparableConvWeights(32, 64),
+          initSeparableConvWeights(64, 64),
+          initSeparableConvWeights(64, 64),
+          initSeparableConvWeights(64, 64),
+          initSeparableConvWeights(64, 128),
+          initSeparableConvWeights(128, 128),
+          initSeparableConvWeights(128, 128),
+          initSeparableConvWeights(128, 128),
+          initSeparableConvWeights(128, 256),
+          initSeparableConvWeights(256, 256),
+          initSeparableConvWeights(256, 256),
+          initSeparableConvWeights(256, 256)
+        ])
+
+      const fc = flatten(
+        [
+          initializer.apply([1, 1, isTiny ? 128 : 256, 136]),
+          // bias
+          tf.zeros([136])
+        ]
+          .map(toDataArray)
+      )
+      return new Float32Array(separableConvWeights.concat(fc))
+    }
+
+    function save(isTiny = false) {
+      const initialWeights = initWeights(
+        tf.initializers.glorotNormal(),
+        isTiny
+      )
+      saveAs(new Blob([initialWeights]), `initial_glorot.weights`)
+    }
+
+  </script>
+</body>
+</html>
\ No newline at end of file

tools/train/faceLandmarks/js/indexedDb.js

+const  indexedDB = window.indexedDB || window.mozIndexedDB || window.webkitIndexedDB || window.msIndexedDB || window.shimIndexedDB
+
+let jsonsDb = null
+let jpgsDb = null
+
+const openJsons = indexedDB.open('jsons', 1)
+const openJpgs = indexedDB.open('jpgs', 1)
+
+openJsons.onupgradeneeded = function() {
+  jsonsDb = openJsons.result
+  jsonsDb.createObjectStore('jsons', { keyPath: 'id' })
+}
+openJpgs.onupgradeneeded = function() {
+  jpgsDb = openJpgs.result
+  jpgsDb.createObjectStore('jpgs', { keyPath: 'id' })
+}
+openJsons.onsuccess = function() {
+  console.log('connected to jsons')
+  jsonsDb = openJsons.result
+}
+openJpgs.onsuccess = function() {
+  console.log('connected to jpgs')
+  jpgsDb = openJpgs.result
+}
+
+function putReq(store, obj) {
+  return new Promise((res, rej) => {
+    const req = store.put(obj)
+    req.onsuccess = res
+    req.onerror = rej
+  })
+}
+
+function getReq(store, id, throwIfNoResult = true) {
+  return new Promise((res, rej) => {
+    const req = store.get(id)
+    req.onsuccess = () => {
+      if (!req.result && throwIfNoResult) {
+        return rej(`no result for id: ${id}`)
+      }
+      res(req.result)
+    }
+    req.onerror = rej
+  })
+}
+
+function existsReq(store, id) {
+  return getReq(store, id, false)
+}
+
+async function getNotFound(store, ids) {
+  return (await Promise.all(ids.map(async id => ({ id, exists: await existsReq(store, id) }))))
+    .filter(({ exists }) => !exists)
+    .map(({ id }) => id)
+}
+
+async function getNotFoundPts(ids) {
+  const store = jsonsDb.transaction('jsons', 'readonly').objectStore('jsons')
+  return getNotFound(store, ids)
+}
+
+async function getNotFoundJpegs(ids) {
+  const store = jpgsDb.transaction('jpgs', 'readonly').objectStore('jpgs')
+  return getNotFound(store, ids)
+}
+
+async function persistPts(ptsById, overwrite = false) {
+  const store = jsonsDb.transaction('jsons', 'readwrite').objectStore('jsons')
+  for (let i = 0; i < ptsById.length; i++) {
+    const { id, pts } = ptsById[i]
+    if (!await existsReq(store, id)) {
+      console.log('persistPts - inserting %s', id)
+      await putReq(store, { id, pts })
+    }
+  }
+}
+
+function getPts(ids) {
+  const store = jsonsDb.transaction('jsons', 'readonly').objectStore('jsons')
+  return Promise.all(ids.map(id => getReq(store, id)))
+}
+
+async function persistJpgs(jpgsById, overwrite = false) {
+  const store = jpgsDb.transaction('jpgs', 'readwrite').objectStore('jpgs')
+  for (let i = 0; i < jpgsById.length; i++) {
+    const { id, blob } = jpgsById[i]
+    if (!await existsReq(store, id)) {
+      console.log('persistJpgs - inserting %s', id)
+      await putReq(store, { id, blob })
+    }
+  }
+}
+
+function getJpgs(ids) {
+  const store = jpgsDb.transaction('jpgs', 'readonly').objectStore('jpgs')
+  return Promise.all(ids.map(id => getReq(store, id)))
+}
+

tools/train/faceLandmarks/js/printIndexDbSizes.js

+// https://gist.github.com/tralves/9e5de2bd9f582007a52708d7d4209865
+
+var getTableSize = function(db, dbName){
+  return new Promise((resolve,reject) => {
+    if (db == null) {
+      return reject();
+    }
+    var size = 0;
+    db = event.target.result;
+    var transaction = db.transaction([dbName])
+      .objectStore(dbName)
+      .openCursor();
+
+    transaction.onsuccess = function(event){
+        var cursor = event.target.result;
+        if(cursor){
+            var storedObject = cursor.value;
+            var json = JSON.stringify(storedObject);
+            size += json.length;
+            cursor.continue();
+        }
+        else{
+          resolve(size);
+        }
+    }.bind(this);
+    transaction.onerror = function(err){
+        reject("error in " + dbName + ": " + err);
+    }
+  });
+};
+
+var getDatabaseSize = function (dbName) {
+  var request = indexedDB.open(dbName);
+  var db;
+  var dbSize = 0;
+  request.onerror = function(event) {
+    alert("Why didn't you allow my web app to use IndexedDB?!");
+  };
+  request.onsuccess = function(event) {
+    db = event.target.result;
+    var tableNames = [ ...db.objectStoreNames ];
+    (function(tableNames, db) {
+      var tableSizeGetters = tableNames
+        .reduce( (acc, tableName) => {
+          acc.push( getTableSize(db, tableName) );
+          return acc;
+        }, []);
+
+      Promise.all(tableSizeGetters)
+        .then(sizes => {
+          console.log('--------- ' + db.name + ' -------------');
+          tableNames.forEach( (tableName,i) => {
+            console.log(" - " + tableName + "\t: " + humanReadableSize(sizes[i]));
+          });
+          var total = sizes.reduce(function(acc, val) {
+            return acc + val;
+          }, 0);
+
+          console.log("TOTAL: " + humanReadableSize(total))
+        });
+      })(tableNames, db);
+  };
+};
+
+var humanReadableSize = function (bytes) {
+  var thresh = 1024;
+  if(Math.abs(bytes) < thresh) {
+    return bytes + ' B';
+  }
+  var units = ['KB','MB','GB','TB','PB','EB','ZB','YB'];
+  var u = -1;
+  do {
+    bytes /= thresh;
+    ++u;
+  } while(Math.abs(bytes) >= thresh && u < units.length - 1);
+  return bytes.toFixed(1)+' '+units[u];
+}
+
+getDatabaseSize('jsons')
+getDatabaseSize('jpgs')
\ No newline at end of file

tools/train/faceLandmarks/js/randomCrop.js

+function randomCrop(img, pts) {
+  const xCoords = pts.map(pt => pt.x)
+  const yCoords = pts.map(pt => pt.y)
+  const minX = xCoords.reduce((x, min) => x < min ? x : min, img.width)
+  const minY = yCoords.reduce((y, min) => y < min ? y : min, img.height)
+  const maxX = xCoords.reduce((x, max) => x < max ? max : x, 0)
+  const maxY = yCoords.reduce((y, max) => y < max ? max : y, 0)
+
+  const x0 = Math.random() * minX
+  const y0 = Math.random() * minY
+
+  const x1 = (Math.random() * Math.abs(img.width - maxX)) + maxX
+  const y1 = (Math.random() * Math.abs(img.height - maxY)) + maxY
+
+  const targetCanvas = faceapi.createCanvas({ width: x1 - x0, height: y1 - y0 })
+  const inputCanvas = img instanceof HTMLCanvasElement ? img : faceapi.createCanvasFromMedia(img)
+  const region = faceapi.getContext2dOrThrow(inputCanvas)
+    .getImageData(x0, y0, targetCanvas.width, targetCanvas.height)
+  faceapi.getContext2dOrThrow(targetCanvas).putImageData(region, 0, 0)
+
+  return {
+    croppedImage: targetCanvas,
+    shiftedPts: pts.map(pt => new faceapi.Point(pt.x - x0, pt.y - y0))
+  }
+}
\ No newline at end of file

tools/train/faceLandmarks/js/synchronizeDb.js

+async function fillPtsDb(ids) {
+  const numFetchPerIteration = 4000
+
+  for (let i = 0; i < Math.floor(ids.length / numFetchPerIteration) + 1; i++) {
+    const ptsById = await Promise.all(
+      ids.slice(i * numFetchPerIteration, (i + 1) * numFetchPerIteration)
+        .map(async id => ({
+          id,
+          pts: await (await fetch(`${id}.json`)).json()
+        }))
+    )
+
+    console.log('persistPts')
+    console.time('persistPts')
+    await persistPts(ptsById)
+    console.timeEnd('persistPts')
+  }
+}
+
+async function fillJpgsDb(ids) {
+  const numFetchPerIteration = 4000
+
+  for (let i = 0; i < Math.floor(ids.length / numFetchPerIteration) + 1; i++) {
+    const jpgsById = await Promise.all(
+      ids.slice(i * numFetchPerIteration, (i + 1) * numFetchPerIteration)
+        .map(async id => ({
+          id,
+          blob: await fetchImage(`${id}.jpg`)
+        }))
+    )
+    console.log('persistJpgs')
+    console.time('persistJpgs')
+    await persistJpgs(jpgsById)
+    console.timeEnd('persistJpgs')
+  }
+}
\ No newline at end of file

tools/train/faceLandmarks/test.html

+<!DOCTYPE html>
+<html>
+<head>
+  <script src="face-api.js"></script>
+  <script src="commons.js"></script>
+  <script src="js/randomCrop.js"></script>
+  <script src="js/indexedDb.js"></script>
+  <script src="js/synchronizeDb.js"></script>
+  <link rel="stylesheet" href="styles.css">
+  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
+  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
+  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
+</head>
+<body>
+  <div id="result" style="position: relative" class="margin">
+    <img id="inputImg" src="" style="max-width: 400px;" />
+    <canvas id="overlay" />
+  </div>
+  <script>
+
+    const net = new faceapi.FaceLandmark68LargeNet()
+    const modelCheckpoint = 'tmp/densenet4/checkpoints/landmarks_epoch46_lr00001_12_lr000001_18.weights'
+    const crops = 4
+
+    async function loadNetWeights(uri) {
+      return new Float32Array(await (await fetch(uri)).arrayBuffer())
+    }
+
+    async function load() {
+      //await net.load('./')
+      const weights = await loadNetWeights(modelCheckpoint)
+      await net.load(weights)
+      console.log('model loaded')
+    }
+
+    load()
+
+    function pointsToFlatArray(pts) {
+      return pts.map(pt => [pt.x, pt.y]).reduce((flat, arr) => flat.concat(arr))
+    }
+
+    async function showResults(landmarks) {
+      const inputImgEl = $('#inputImg').get(0)
+      const { width, height } = inputImgEl
+      const canvas = $('#overlay').get(0)
+      canvas.width = width
+      canvas.height = height
+
+      faceapi.drawLandmarks('overlay', landmarks, { drawLines: true })
+    }
+
+    async function testAll() {
+      window.testIds = (await fetch('/face_landmarks_test_ids').then(res => res.json()))
+      test()
+    }
+
+    async function testSidewaysX(useTrainSet = false) {
+      window.testIds = (await fetch(useTrainSet ? '/sideways_x_train' : '/sideways_x_test').then(res => res.json()))
+      test()
+    }
+
+    async function testStrongSidewaysX(useTrainSet = false) {
+      window.testIds = (await fetch(useTrainSet ? '/strong_sideways_x_train' : '/strong_sideways_x_test').then(res => res.json()))
+      test()
+    }
+
+    async function testStrongSidewaysY(useTrainSet = false) {
+      window.testIds = (await fetch(useTrainSet ? '/strong_sideways_y_train' : '/strong_sideways_y_test').then(res => res.json()))
+      test()
+    }
+
+    async function test() {
+      let totalError = 0
+
+      for (let i = 0; i < window.testIds.length; i++) {
+
+        const dataId = window.testIds[i]
+        let originalImage = (await getJpgs([dataId]))[0]
+        originalImage = await faceapi.bufferToImage(originalImage.blob)
+        const groundTruthPts = (await getPts([dataId])).map(res => res.pts)[0]
+
+        for (let c = 0; c < (crops || 1); c++) {
+
+          console.log(i, c, i / window.testIds.length)
+
+          const { croppedImage, shiftedPts } = crops > 0
+            ? randomCrop(originalImage, groundTruthPts)
+            : { croppedImage: originalImage, shiftedPts: groundTruthPts }
+
+          const squaredImg = faceapi.imageToSquare(croppedImage, 112, true)
+
+          const landmarks = (await net.detectLandmarks(squaredImg)).forSize(croppedImage.width, croppedImage.height)
+
+          const dist = faceapi.euclideanDistance(
+            pointsToFlatArray(landmarks.getPositions()),
+            pointsToFlatArray(shiftedPts)
+          )
+
+          const distByCoordinate = dist / 136
+          totalError += distByCoordinate
+
+          if (window.debug) {
+            const inputImgEl = $('#inputImg').get(0)
+            inputImgEl.src = croppedImage instanceof HTMLImageElement ? croppedImage.src : croppedImage.toDataURL()
+            inputImgEl.onload = () => showResults(landmarks)
+
+            for (let i = 0; i < 30; i++)
+              await faceapi.tf.nextFrame()
+          }
+        }
+      }
+
+      console.log('done...')
+      console.log('test set size:', window.testIds.length)
+      console.log('total error:', totalError / (crops || 1))
+      console.log('average error:', totalError / ((crops || 1) * window.testIds.length))
+    }
+
+
+
+  </script>
+</body>
+</html>
\ No newline at end of file

tools/train/faceLandmarks/train.html

 <!DOCTYPE html>
 <html>
 <head>
-  <link rel="stylesheet" href="styles.css">
-  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
-  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
-  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
   <script src="face-api.js"></script>
   <script src="commons.js"></script>
-  <script src="faceLandmarksUi.js"></script>
-  <script src="faceLandmarksTrain.js"></script>
+  <script src="js/randomCrop.js"></script>
+  <script src="js/indexedDb.js"></script>
+  <script src="js/synchronizeDb.js"></script>
   <script src="FileSaver.js"></script>
-  <script src="trainUtils.js"></script>
 </head>
 <body>
-  <div id="navbar"></div>
-  <div class="center-content page-container">
-    <div>
-      <div class="progress" id="loader">
-        <div class="indeterminate"></div>
-      </div>
-      <div class="row side-by-side">
-        <div class="center-content">
-          <div id="faceContainer"></div>
-          <div id="selectList"></div>
-        </div>
-      </div>
-      <p>
-        <input type="checkbox" id="drawLinesCheckbox" checked="checked" onchange="onChangeDrawLines(event)" />
-        <label for="drawLinesCheckbox">Draw Lines</label>
-      </p>
-      <div class="row">
-        <label for="imgByNr">Enter image NR: </label>
-        <input id="imgByNr" type="text" class="bold">
-      </div>
-    </div>
-  </div>
+  <div id="container"></div>
 
   <script>
     tf = faceapi.tf
-    window.saveEveryNthIteration = 2
-    window.drawLines = true
-    window.nets = []
-    window.landmarks = []
-
-    window.trainSteps = 100
-    window.learningRate = 0.01
-    //window.optimizer = tf.train.sgd(learningRate)
-    window.optimizer = tf.train.adam(0.001, 0.9, 0.999, 1e-8)
-
-    function lossFunction(labels, out) {
-      return tf.losses.meanSquaredError(labels, out)
+
+    // uri to weights file of last checkpoint
+    window.net = new faceapi.FaceLandmark68TinyNet()
+    const modelCheckpoint = 'tmp/face_landmark_68_tiny_model_lr00001_19.weights'
+    const startEpoch = 20
+
+    const learningRate = 0.0001 // 0.001
+    window.optimizer = tf.train.adam(learningRate, 0.9, 0.999, 1e-8)
+
+    window.saveEveryNthSample = Infinity
+
+    window.withRandomCrop = true
+
+    window.batchSize = 12
+    //window.batchSize = 32
+
+    window.lossValues = []
+
+    window.iterDelay = 0
+    window.withLogging = true
+
+    const log = (str, ...args) => console.log(`[${[(new Date()).toTimeString().substr(0, 8)]}] ${str || ''}`, ...args)
+
+    function saveWeights(net, filename = 'train_tmp') {
+      saveAs(new Blob([net.serializeParams()]), filename)
+    }
+
+    async function loadNetWeights(uri) {
+      return new Float32Array(await (await fetch(uri)).arrayBuffer())
+    }
+
+    async function delay(ms) {
+      return new Promise(res => setTimeout(res, ms))
     }
 
-    async function run() {
-      window.trainData = await getTrainData()
-      window.trainNet.variable()
+    async function fetchTrainDataIds() {
+      return fetch('/face_landmarks_train_ids').then(res => res.json())
+    }
+
+    async function load() {
+      window.trainIds = (await fetchTrainDataIds())
+      const weights = await loadNetWeights(modelCheckpoint)
+      await window.net.load(weights)
+      window.net.variable()
+    }
+
+    async function onEpochDone(epoch) {
+      saveWeights(window.net, `face_landmark_68_tiny_model_lr00001_${epoch}.weights`)
 
-      return
+      const loss = window.lossValues[epoch]
+      saveAs(new Blob([JSON.stringify({ loss, avgLoss: loss / window.trainIds.length })]), `face_landmark_68_tiny_model_lr00001_${epoch}.json`)
 
-      await train()
     }
 
-    async function train(batchSize = 1) {
-      for (let i = 0; i < trainSteps; i++) {
-        console.log('step', i)
-        const batchCreators = createBatchCreators(shuffle(window.trainData), batchSize)
-        let ts = Date.now()
-        await trainStep(batchCreators)
-        ts = Date.now() - ts
-        console.log('step %s done (%s ms)', i, ts)
-        if (((i + 1) % saveEveryNthIteration) === 0) {
-          //saveWeights(window.trainNet, 'landmark_trained_weights_' + idx + '.weights')
+    async function train() {
+      await load()
+
+      for (let epoch = startEpoch; epoch < Infinity; epoch++) {
+
+        if (epoch !== startEpoch) {
+          // ugly hack to wait for loss datas for that epoch to be resolved
+          setTimeout(() => onEpochDone(epoch - 1), 10000)
+        }
+        window.lossValues[epoch] = 0
+
+        const shuffledInputs = faceapi.shuffleArray(window.trainIds)
+
+        for (let dataIdx = 0; dataIdx < shuffledInputs.length; dataIdx += window.batchSize) {
+          const tsIter = Date.now()
+
+          const ids = shuffledInputs.slice(dataIdx, dataIdx + window.batchSize)
+
+          // fetch image and ground truth bounding boxes
+          let tsFetch = Date.now()
+          let tsFetchPts = Date.now()
+
+          const allPts = (await getPts(ids)).map(res => res.pts)
+
+          tsFetchPts = Date.now() - tsFetchPts
+
+          let tsFetchJpgs = Date.now()
+
+          const allJpgBlobs = (await getJpgs(ids)).map(res => res.blob)
+
+          tsFetchJpgs = Date.now() - tsFetchJpgs
+
+          let tsBufferToImage = Date.now()
+
+          const allJpgs = await Promise.all(
+            allJpgBlobs.map(blob => faceapi.bufferToImage(blob))
+          )
+
+          tsBufferToImage = Date.now() - tsBufferToImage
+
+          tsFetch = Date.now() - tsFetch
+
+          const bImages = []
+          const bPts = []
+
+          for (let batchIdx = 0; batchIdx < ids.length; batchIdx += 1) {
+
+            const pts = allPts[batchIdx]
+            const img = allJpgs[batchIdx]
+
+            const { croppedImage, shiftedPts } = window.withRandomCrop
+              ? randomCrop(img, pts)
+              : { croppedImage: img, shiftedPts: pts }
+
+            const squaredImg = faceapi.imageToSquare(croppedImage, 112, true)
+
+            const reshapedDimensions = faceapi.computeReshapedDimensions(croppedImage, 112)
+
+            const pad = Math.abs(reshapedDimensions.width - reshapedDimensions.height) /  (2 * 112)
+            const padX = reshapedDimensions.width < reshapedDimensions.height ? pad : 0
+            const padY = reshapedDimensions.height < reshapedDimensions.width ? pad : 0
+
+            const groundTruthLandmarks = shiftedPts.map(pt => new faceapi.Point(
+              padX + (pt.x / croppedImage.width) * (reshapedDimensions.width / 112),
+              padY + (pt.y / croppedImage.height)* (reshapedDimensions.height / 112)
+            ))
+
+            bImages.push(squaredImg)
+            bPts.push(groundTruthLandmarks)
+          }
+
+          const netInput = await faceapi.toNetInput(bImages)
+
+          let tsBackward = Date.now()
+          let tsForward = Date.now()
+
+          const loss = optimizer.minimize(() => {
+            const out = window.net.runNet(netInput)
+            tsForward = Date.now() - tsForward
+            tsBackward = Date.now()
+            const landmarksTensor = tf.tensor2d(
+              bPts
+                .reduce((flat, arr) => flat.concat(arr))
+                .map(pt => [pt.x, pt.y])
+                .reduce((flat, arr) => flat.concat(arr)),
+              [ids.length, 136]
+            )
+
+            const loss = tf.losses.meanSquaredError(
+              landmarksTensor,
+              out,
+              tf.Reduction.MEAN
+            )
+
+            return loss
+          }, true)
+          tsBackward = Date.now() - tsBackward
+
+          // start next iteration without waiting for loss data
+
+          loss.data().then(data => {
+            const lossValue = data[0]
+            window.lossValues[epoch] += lossValue
+            window.withLogging && log(`epoch ${epoch}, dataIdx ${dataIdx} - loss: ${lossValue}, ( ${window.lossValues[epoch]})`)
+            loss.dispose()
+          })
+
+          window.withLogging && log(`epoch ${epoch}, dataIdx ${dataIdx} - forward: ${tsForward} ms, backprop: ${tsBackward} ms, iter: ${Date.now() - tsIter} ms`)
+          if (window.logsilly)  {
+            log(`fetch: ${tsFetch} ms, pts: ${tsFetchPts} ms, jpgs: ${tsFetchJpgs} ms, bufferToImage: ${tsBufferToImage} ms`)
+          }
+          if (window.iterDelay) {
+            await delay(window.iterDelay)
+          } else {
+            await tf.nextFrame()
+          }
         }
       }
     }
 
   </script>
-
 </body>
 </html>
\ No newline at end of file

tools/train/faceLandmarks/train_debug.html

+<!DOCTYPE html>
+<html>
+<head>
+  <script src="face-api.js"></script>
+  <script src="commons.js"></script>
+  <script src="js/randomCrop.js"></script>
+  <script src="js/indexedDb.js"></script>
+  <script src="FileSaver.js"></script>
+</head>
+<body>
+  <div id="container"></div>
+
+  <script>
+    tf = faceapi.tf
+
+    // uri to weights file of last checkpoint
+    window.net = new faceapi.FaceLandmark68LargeNet()
+    const modelCheckpoint = '/tmp/initial_glorot.weights'
+    const startEpoch = 0
+
+    const learningRate = 0.001 // 0.001
+    window.optimizer = tf.train.adam(learningRate, 0.9, 0.999, 1e-8)
+
+    window.saveEveryNthSample = 50000000
+
+    window.withRandomCrop = false
+
+    window.batchSize = 12
+
+    window.lossValue = 0
+
+    window.drawResults = true
+
+    const log = (str, ...args) => console.log(`[${[(new Date()).toTimeString().substr(0, 8)]}] ${str || ''}`, ...args)
+
+    function saveWeights(net, filename = 'train_tmp') {
+      saveAs(new Blob([net.serializeParams()]), filename)
+    }
+
+    async function loadNetWeights(uri) {
+      return new Float32Array(await (await fetch(uri)).arrayBuffer())
+    }
+
+    async function fetchTrainDataFilenames() {
+      return fetch('/face_landmarks_train_ids').then(res => res.json())
+    }
+
+    async function load() {
+      window.ptsFiles = await fetchTrainDataFilenames()
+
+      window.ptsFiles = window.ptsFiles.slice(1000, 1000 + window.batchSize)
+      //window.ptsFiles = window.ptsFiles.slice(7, 8)
+
+      const weights = await loadNetWeights(modelCheckpoint)
+      await window.net.load(weights)
+      window.net.variable()
+    }
+
+    function randomCrop(img, pts) {
+      const xCoords = pts.map(pt => pt.x)
+      const yCoords = pts.map(pt => pt.y)
+      const minX = xCoords.reduce((x, min) => x < min ? x : min, img.width)
+      const minY = yCoords.reduce((y, min) => y < min ? y : min, img.height)
+      const maxX = xCoords.reduce((x, max) => x < max ? max : x, 0)
+      const maxY = yCoords.reduce((y, max) => y < max ? max : y, 0)
+
+      const x0 = Math.random() * minX
+      const y0 = Math.random() * minY
+
+      const x1 = (Math.random() * Math.abs(img.width - maxX)) + maxX
+      const y1 = (Math.random() * Math.abs(img.height - maxY)) + maxY
+
+      const targetCanvas = faceapi.createCanvas({ width: x1 - x0, height: y1 - y0 })
+      const inputCanvas = img instanceof HTMLCanvasElement ? img : faceapi.createCanvasFromMedia(img)
+      const region = faceapi.getContext2dOrThrow(inputCanvas)
+        .getImageData(x0, y0, targetCanvas.width, targetCanvas.height)
+      faceapi.getContext2dOrThrow(targetCanvas).putImageData(region, 0, 0)
+
+      return {
+        croppedImage: targetCanvas,
+        shiftedPts: pts.map(pt => new faceapi.Point(pt.x - x0, pt.y - y0))
+      }
+    }
+
+    async function train() {
+      await load()
+
+      for (let epoch = startEpoch; epoch < Infinity; epoch++) {
+
+        if (epoch !== startEpoch) {
+          //saveWeights(window.net, `landmarks_epoch${epoch - 1}.weights`)
+          //saveAs(new Blob([JSON.stringify({ loss: window.lossValue, avgLoss: window.lossValue / 20000 })]), `landmarks_epoch${epoch - 1}.json`)
+          window.lossValue = 0
+        }
+
+        const shuffledInputs = window.drawResults ? window.ptsFiles : faceapi.shuffleArray(window.ptsFiles)
+
+        for (let dataIdx = 0; dataIdx < shuffledInputs.length; dataIdx += window.batchSize) {
+
+          const ids = shuffledInputs.slice(dataIdx, dataIdx + window.batchSize)
+
+          // fetch image and ground truth bounding boxes
+          console.time('getPts')
+          const allPts = (await getPts(ids)).map(res => res.pts)
+          console.timeEnd('getPts')
+          console.time('getJpgs')
+          const allJpgs = await Promise.all(
+            (await getJpgs(ids)).map(res => res.blob).map(blob => faceapi.bufferToImage(blob))
+          )
+          console.timeEnd('getJpgs')
+
+          const bSquaredImages = []
+          const bCroppedImages = []
+          const bImages = []
+          const bPts = []
+
+          for (let batchIdx = 0; batchIdx < ids.length; batchIdx += 1) {
+
+            const pts = allPts[batchIdx]
+            const img = allJpgs[batchIdx]
+
+            const { croppedImage, shiftedPts } = window.withRandomCrop
+              ? randomCrop(img, pts)
+              : { croppedImage: img, shiftedPts: pts }
+
+            const squaredImg = faceapi.imageToSquare(croppedImage, 112, true)
+
+            const reshapedDimensions = faceapi.computeReshapedDimensions(croppedImage, 112)
+
+            const pad = Math.abs(reshapedDimensions.width - reshapedDimensions.height) /  (2 * 112)
+            const padX = reshapedDimensions.width < reshapedDimensions.height ? pad : 0
+            const padY = reshapedDimensions.height < reshapedDimensions.width ? pad : 0
+
+            const groundTruthLandmarks = shiftedPts.map(pt => new faceapi.Point(
+              padX + (pt.x / croppedImage.width) * (reshapedDimensions.width / 112),
+              padY + (pt.y / croppedImage.height)* (reshapedDimensions.height / 112)
+            ))
+
+            bSquaredImages.push(squaredImg)
+            bCroppedImages.push(croppedImage)
+            bImages.push(squaredImg)
+            bPts.push(groundTruthLandmarks)
+          }
+
+          const netInput = await faceapi.toNetInput(bImages)
+
+          const ts = Date.now()
+
+          const loss = optimizer.minimize(() => {
+            console.time('runNet')
+            const out = window.net.runNet(netInput)
+            console.timeEnd('runNet')
+
+            if (window.drawResults) {
+              // debug
+              const landmarkTensors = tf.unstack(out)
+              landmarkTensors.forEach((t, i) => {
+                const croppedImage = bSquaredImages[i]
+
+
+                const landmarksArray = Array.from(t.dataSync())
+                const xCoords = landmarksArray.filter((_, i) => faceapi.isEven(i))
+                const yCoords = landmarksArray.filter((_, i) => !faceapi.isEven(i))
+ /*
+                const landmarksArray = bPts[i]
+                const xCoords = landmarksArray.map(pt => pt.x)
+                const yCoords = landmarksArray.map(pt => pt.y)
+*/
+                const marks = new faceapi.FaceLandmarks68(
+                  Array(68).fill(0).map((_, i) => new faceapi.Point(xCoords[i], yCoords[i])),
+                  croppedImage
+                )
+
+                const canvas = croppedImage instanceof HTMLCanvasElement ? croppedImage : faceapi.createCanvasFromMedia(croppedImage)
+                faceapi.drawLandmarks(canvas, marks, { drawLines: true })
+                const container = document.getElementById('container')
+                if (container.children[i]) {
+                  const oldChild = container.children[i]
+                  container.insertBefore(canvas, oldChild)
+                  container.removeChild(oldChild)
+                } else {
+                  container.appendChild(canvas)
+                }
+              })
+              // debug
+            }
+
+            const landmarksTensor = tf.tensor2d(
+              bPts
+                .reduce((flat, arr) => flat.concat(arr))
+                .map(pt => [pt.x, pt.y])
+                .reduce((flat, arr) => flat.concat(arr)),
+              [window.batchSize, 136]
+            )
+
+            const loss = tf.losses.meanSquaredError(
+              landmarksTensor,
+              out,
+              tf.Reduction.MEAN
+            )
+
+            //return tf.sum(tf.abs(tf.sub(landmarksTensor, out)))
+            return loss
+          }, true)
+
+          // start next iteration without waiting for loss data
+          loss.data().then(data => {
+            const lossValue = data[0]
+            window.lossValue += lossValue
+            log(`epoch ${epoch}, dataIdx ${dataIdx} - loss: ${lossValue}`)
+            loss.dispose()
+          })
+
+          if (window.iterDelay) {
+            await delay(window.iterDelay)
+          } else {
+            await tf.nextFrame()
+          }
+        }
+      }
+    }
+
+  </script>
+</body>
+</html>
\ No newline at end of file

tools/train/faceLandmarks/verify.html

+<!DOCTYPE html>
+<html>
+<head>
+  <script src="face-api.js"></script>
+  <script src="commons.js"></script>
+  <script src="js/randomCrop.js"></script>
+  <link rel="stylesheet" href="styles.css">
+  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
+  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
+  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
+</head>
+<body>
+  <div id="navbar"></div>
+  <div class="center-content page-container">
+    <div class="progress" id="loader">
+      <div class="indeterminate"></div>
+    </div>
+    <div id="results" class="side-by-side">
+      <div id="result0" style="position: relative" class="margin">
+        <img id="inputImg0" src="" style="max-width: 400px;" />
+        <canvas class="overlay" id="overlay0" />
+      </div>
+      <div id="result1" style="position: relative" class="margin">
+        <img id="inputImg1" src="" style="max-width: 400px;" />
+        <canvas class="overlay" id="overlay1" />
+      </div>
+      <div id="result2" style="position: relative" class="margin">
+        <img id="inputImg2" src="" style="max-width: 400px;" />
+        <canvas class="overlay" id="overlay2" />
+      </div>
+    </div>
+    <div class="row">
+      <label for="imgByNr">Enter image NR: </label>
+      <input id="imgByNr" type="text" class="bold" value="0"/>
+    </div>
+  </div>
+
+  <script>
+
+    const modelCheckpoints = [
+      'tmp/face_landmark_68_tiny_model_lr00001_0.weights',
+      'tmp/face_landmark_68_tiny_model_lr00001_13.weights',
+      'tmp/face_landmark_68_tiny_model_lr00001_24.weights'
+    ]
+    let originalImage = null
+
+    function cloneImage(img) {
+      var image = document.createElement("img")
+      image.src = img.src
+      return image
+    }
+
+    function getInputValue(input) {
+      input = input || $('#imgByNr').get(0)
+      return input.value || 0
+    }
+
+    function onKeyDown(e) {
+      e.target.value = (
+        parseInt(e.target.value) + (e.keyCode === 38 ? 1 : (e.keyCode === 40 ? -1 : 0))
+      ) || getInputValue(e.target)
+
+
+      const imgUri = `${window.testIds[e.target.value]}.jpg`
+
+      console.log(imgUri)
+      onSelectionChanged(imgUri)
+    }
+
+    async function updateResults(_landmarks) {
+      const inputImgEl = $('#inputImg0').get(0)
+      const { width, height } = inputImgEl
+
+      for (let i = 0; i < window.nets.length; i++) {
+        const net = window.nets[i]
+
+        const canvas = $(`#overlay${i}`).get(0)
+        canvas.width = width
+        canvas.height = height
+        console.time('detectLandmarks')
+        const landmarks = await net.detectLandmarks(inputImgEl)
+        console.timeEnd('detectLandmarks')
+        faceapi.drawLandmarks(`overlay${i}`, landmarks.forSize(width, height), { drawLines: true })
+      }
+
+    }
+
+    async function onSelectionChanged(imgUri) {
+      const img = await faceapi.bufferToImage(await fetchImage(imgUri))
+
+      window.nets.forEach(async (_, i) => {
+        $(`#inputImg${i}`).get(0).src = img.src
+      })
+      originalImage = cloneImage(img)
+      updateResults()
+    }
+
+    async function applyRandomCrop() {
+      const dataId = window.testIds[getInputValue()]
+      const pts = await (await fetch(`${dataId}.json`)).json()
+
+      const img = cloneImage(originalImage)
+
+      const { croppedImage, shiftedPts } = randomCrop(img, pts)
+
+      img.src = croppedImage.toDataURL()
+      $(`#result${0}`).get(0).removeChild($(`#inputImg${0}`).get(0))
+      $(`#result${0}`).get(0).appendChild(img)
+      window.nets.slice(1).forEach((_, i) => {
+        const im = cloneImage(img)
+        im.id = `#inputImg${i + 1}`
+        $(`#result${i + 1}`).get(0).removeChild($(`#inputImg${i + 1}`).get(0))
+        $(`#result${i + 1}`).get(0).appendChild(im)
+      })
+      const { width, height } = croppedImage
+      img.onload = () => updateResults(
+        new faceapi.FaceLandmarks68(
+          shiftedPts.map(pt =>
+            new faceapi.Point(
+              pt.x / width,
+              pt.y / height
+            )
+          ),
+          { width, height }
+        )
+      )
+    }
+
+    async function loadNetWeights(uri) {
+      return new Float32Array(await (await fetch(uri)).arrayBuffer())
+    }
+
+    async function run() {
+      $('#imgByNr').keydown(onKeyDown)
+
+      //faceapi.loadFaceLandmarkModel('./')
+
+      //window.trainIds = (await fetch('/face_landmarks_train_ids').then(res => res.json()))
+      //window.testIds = (await fetch('/face_landmarks_test_ids').then(res => res.json()))
+      window.testIds = (await fetch('/strong_sideways_test').then(res => res.json()))
+
+      window.nets = []
+      //window.nets.push(faceapi)
+      //const net = new faceapi.FaceLandmark68TinyNet()
+      //await net.load('./')
+      //window.nets.push(net)
+      modelCheckpoints.forEach(async checkpoint => {
+        const net = new faceapi.FaceLandmark68TinyNet()
+        const weights = await loadNetWeights(checkpoint)
+        await net.load(weights)
+        window.nets.push(net)
+      })
+
+      $('#loader').hide()
+    }
+
+    $(document).ready(() => run())
+
+  </script>
+</body>
+</html>
\ No newline at end of file

tools/train/serveFaceLandmarks.js

@@ -2,6 +2,7 @@ require('./faceLandmarks/.env')
 
 const express = require('express')
 const path = require('path')
+const fs = require('fs')
 
 const app = express()
 
@@ -14,9 +15,72 @@ app.use(express.static(path.join(__dirname, '../../weights')))
 app.use(express.static(path.join(__dirname, '../../dist')))
 
 const trainDataPath = path.resolve(process.env.TRAIN_DATA_PATH)
-app.use(express.static(trainDataPath))
 
-app.get('/', (req, res) => res.redirect('/face_landmarks'))
-app.get('/face_landmarks', (req, res) => res.sendFile(path.join(publicDir, 'train.html')))
+const pngPath = path.join(trainDataPath, 'png')
+const jpgPath = path.join(trainDataPath, 'jpg')
+const groundTruthPath = path.join(trainDataPath, 'pts')
 
-app.listen(3000, () => console.log('Listening on port 3000!'))
\ No newline at end of file
+app.use(express.static(jpgPath))
+app.use(express.static(groundTruthPath))
+
+const trainIds = JSON.parse(fs.readFileSync(path.join(publicDir, './trainData.json')))
+const trainIdsSet = new Set(trainIds)
+
+const testFileIds = fs.readdirSync(groundTruthPath)
+  .map(file => file.replace('.json', ''))
+  .filter(file => !trainIdsSet.has(file))
+
+app.get('/face_landmarks_train_ids', (req, res) => res.status(202).send(trainIds))
+app.get('/face_landmarks_test_ids', (req, res) => res.status(202).send(testFileIds))
+
+const poseAnchors = JSON.parse(fs.readFileSync(path.join(trainDataPath, './pose-anchors.json')))
+
+const sidewaysXTrain = []
+const strongSidewaysXTrain = []
+const strongSidewaysYTrain = []
+const strongSidewaysTrain = []
+const sidewaysXTest = []
+const strongSidewaysXTest = []
+const strongSidewaysYTest = []
+const strongSidewaysTest = []
+
+poseAnchors.forEach((pt, idx) => {
+  const id = `${idx}`
+  if (Math.abs(0.5 - pt.x) > 0.15) {
+    (trainIdsSet.has(id) ? sidewaysXTrain : sidewaysXTest).push(id)
+  }
+  if (Math.abs(0.5 - pt.x) > 0.2) {
+    (trainIdsSet.has(id) ? strongSidewaysXTrain : strongSidewaysXTest).push(id)
+  }
+  if (Math.abs(0.44 - pt.y) > 0.1) {
+    (trainIdsSet.has(id) ? strongSidewaysYTrain : strongSidewaysYTest).push(id)
+  }
+  if (Math.abs(0.5 - pt.x) > 0.2 || Math.abs(0.44 - pt.y) > 0.1) {
+    (trainIdsSet.has(id) ? strongSidewaysTrain : strongSidewaysTest).push(id)
+  }
+})
+
+console.log(sidewaysXTrain.length)
+console.log(strongSidewaysXTrain.length)
+console.log(strongSidewaysYTrain.length)
+console.log(strongSidewaysTrain.length)
+console.log(sidewaysXTest.length)
+console.log(strongSidewaysXTest.length)
+console.log(strongSidewaysYTest.length)
+console.log(strongSidewaysTrain.length)
+
+app.get('/sideways_x_train', (req, res) => res.status(202).send(sidewaysXTrain))
+app.get('/strong_sideways_x_train', (req, res) => res.status(202).send(strongSidewaysXTrain))
+app.get('/strong_sideways_y_train', (req, res) => res.status(202).send(strongSidewaysYTrain))
+app.get('/strong_sideways_train', (req, res) => res.status(202).send(strongSidewaysTrain))
+app.get('/sideways_x_test', (req, res) => res.status(202).send(sidewaysXTest))
+app.get('/strong_sideways_x_test', (req, res) => res.status(202).send(strongSidewaysXTest))
+app.get('/strong_sideways_y_test', (req, res) => res.status(202).send(strongSidewaysYTest))
+app.get('/strong_sideways_test', (req, res) => res.status(202).send(strongSidewaysTest))
+
+app.get('/', (req, res) => res.redirect('/train'))
+app.get('/train', (req, res) => res.sendFile(path.join(publicDir, 'train.html')))
+app.get('/verify', (req, res) => res.sendFile(path.join(publicDir, 'verify.html')))
+app.get('/test', (req, res) => res.sendFile(path.join(publicDir, 'test.html')))
+
+app.listen(8000, () => console.log('Listening on port 8000!'))
\ No newline at end of file

tools/train/serveTinyYolov2.js

-require('./tinyYolov2/.env')
-
-const express = require('express')
-const path = require('path')
-const fs = require('fs')
-
-const app = express()
-
-const publicDir = path.join(__dirname, './tinyYolov2')
-app.use(express.static(publicDir))
-app.use(express.static(path.join(__dirname, './shared')))
-app.use(express.static(path.join(__dirname, './node_modules/file-saver')))
-app.use(express.static(path.join(__dirname, '../../examples/public')))
-app.use(express.static(path.join(__dirname, '../../weights')))
-app.use(express.static(path.join(__dirname, '../../dist')))
-
-const trainDataPath = path.resolve(process.env.TRAIN_DATA_PATH)
-const testDataPath = path.resolve(process.env.TEST_DATA_PATH)
-const imagesPath = path.join(trainDataPath, './final_images')
-const detectionsPath = path.join(trainDataPath, './final_detections')
-app.use(express.static(imagesPath))
-app.use(express.static(detectionsPath))
-app.use(express.static(testDataPath))
-
-const detectionFilenames = fs.readdirSync(detectionsPath)
-const detectionFilenamesMultibox = JSON.parse(fs.readFileSync(path.join(__dirname, './tinyYolov2/multibox.json')))
-
-app.use(express.static(trainDataPath))
-
-app.get('/detection_filenames', (req, res) => res.status(202).send(detectionFilenames))
-app.get('/detection_filenames_multibox', (req, res) => res.status(202).send(detectionFilenamesMultibox))
-app.get('/', (req, res) => res.sendFile(path.join(publicDir, 'train.html')))
-app.get('/verify', (req, res) => res.sendFile(path.join(publicDir, 'verify.html')))
-app.get('/test', (req, res) => res.sendFile(path.join(publicDir, 'test.html')))
-
-app.listen(3000, () => console.log('Listening on port 3000!'))
\ No newline at end of file

tools/train/shared/trainUtils.js

-const log = (str, ...args) => console.log(`[${[(new Date()).toTimeString().substr(0, 8)]}] ${str || ''}`, ...args)
-
-async function promiseSequential(promises) {
-  const curr = promises[0]
-  if (!curr) {
-    return
-  }
-
-  await curr()
-  return promiseSequential(promises.slice(1))
-}
-
-// https://stackoverflow.com/questions/6274339/how-can-i-shuffle-an-array
-function shuffle(a) {
-  var j, x, i;
-  for (i = a.length - 1; i > 0; i--) {
-      j = Math.floor(Math.random() * (i + 1));
-      x = a[i];
-      a[i] = a[j];
-      a[j] = x;
-  }
-  return a;
-}
-
-function saveWeights(net, filename = 'train_tmp') {
-  const binaryWeights = new Float32Array(
-    net.getParamList()
-      .map(({ tensor }) => Array.from(tensor.dataSync()))
-      .reduce((flat, arr) => flat.concat(arr))
-  )
-  saveAs(new Blob([binaryWeights]), filename)
-}
-
-function imageToSquare(img) {
-  const scale = 608 / Math.max(img.height, img.width)
-  const width = scale * img.width
-  const height = scale * img.height
-
-  const canvas1 = faceapi.createCanvasFromMedia(img)
-  const targetCanvas = faceapi.createCanvas({ width: 608, height: 608 })
-  targetCanvas.getContext('2d').putImageData(canvas1.getContext('2d').getImageData(0, 0, width, height), 0, 0)
-  return targetCanvas
-}
-
-function getPaddingsAndReshapedSize(img, inputSize) {
-  const [h, w] = [img.height, img.width]
-  const maxDim = Math.max(h, w)
-
-  const f = inputSize / maxDim
-  const reshapedImgDims = {
-    height: Math.floor(h * f),
-    width: Math.floor(w * f)
-  }
-
-  const paddings = new faceapi.Point(
-    maxDim / img.width,
-    maxDim / img.height
-  )
-
-  return { paddings, reshapedImgDims }
-}
\ No newline at end of file

tools/train/tinyYolov2/initWeights.html

-<!DOCTYPE html>
-<html>
-<head>
-  <script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@0.12.0"> </script>
-  <script src="FileSaver.js"></script>
-</head>
-<body>
-  <script>
-
-    const glorotNormal = tf.initializers.glorotNormal()
-
-    saveWeights()
-
-    function initSeparableConvWeights(inChannels, outChannels) {
-      return {
-        depthwiseFilter: glorotNormal.apply([3, 3, inChannels, 1]),
-        pointwiseFilter: glorotNormal.apply([1, 1, inChannels, outChannels]),
-        bias: tf.zeros([outChannels])
-      }
-    }
-
-    function initTinyYolov2SeparableWeights() {
-
-      const conv0 = initSeparableConvWeights(3, 16)
-      const conv1 = initSeparableConvWeights(16, 32)
-      const conv2 = initSeparableConvWeights(32, 64)
-      const conv3 = initSeparableConvWeights(64, 128)
-      const conv4 = initSeparableConvWeights(128, 256)
-      const conv5 = initSeparableConvWeights(256, 512)
-      const conv6 = initSeparableConvWeights(512, 1024)
-      const conv7 = initSeparableConvWeights(1024, 1024)
-      const conv8 = {
-        filters: glorotNormal.apply([1, 1, 1024, 25]),
-        bias: tf.zeros([25])
-      }
-
-      return {
-        conv0,
-        conv1,
-        conv2,
-        conv3,
-        conv4,
-        conv5,
-        conv6,
-        conv7,
-        conv8
-      }
-    }
-
-    function saveWeights() {
-      const w = initTinyYolov2SeparableWeights()
-      const binaryWeights = new Float32Array(
-        Array(8).fill(0)
-          .map((_, i) => w[`conv${i}`])
-          .map(ps => [ps.depthwiseFilter, ps.pointwiseFilter, ps.bias])
-          .reduce((flat, arr) => flat.concat(arr))
-          .concat([w.conv8.filters, w.conv8.bias])
-          .map(tensor => console.log(tensor) || Array.from(tensor.dataSync()))
-          .reduce((flat, arr) => flat.concat(arr))
-      )
-      saveAs(new Blob([binaryWeights]), 'foo.weights')
-    }
-  </script>
-</body>
-</html>
\ No newline at end of file

tools/train/tinyYolov2/loss.js

-const CELL_SIZE = 32
-
-const getNumCells = inputSize => inputSize / CELL_SIZE
-
-const inverseSigmoid = x => Math.log(x / (1 - x))
-
-function getAnchors() {
-  return window.net.anchors
-}
-
-function squaredSumOverMask(lossTensors, mask) {
-  return tf.tidy(() => tf.sum(tf.square(tf.mul(mask, lossTensors))))
-}
-
-function assignBoxesToAnchors(groundTruthBoxes, reshapedImgDims) {
-
-  const inputSize = Math.max(reshapedImgDims.width, reshapedImgDims.height)
-  const numCells = getNumCells(inputSize)
-
-  return groundTruthBoxes.map(box => {
-    const { left, top, width, height } = box.rescale(reshapedImgDims)
-
-    const ctX = left + (width / 2)
-    const ctY = top + (height / 2)
-
-    const col = Math.floor((ctX / inputSize) * numCells)
-    const row = Math.floor((ctY / inputSize) * numCells)
-
-    const anchorsByIou = getAnchors().map((anchor, idx) => ({
-      idx,
-      iou: faceapi.iou(
-        new faceapi.BoundingBox(0, 0, anchor.x * CELL_SIZE, anchor.y * CELL_SIZE),
-        new faceapi.BoundingBox(0, 0, width, height)
-      )
-    })).sort((a1, a2) => a2.iou - a1.iou)
-
-    const anchor = anchorsByIou[0].idx
-
-    return { row, col, anchor, box }
-  })
-}
-
-function getGroundTruthMask(groundTruthBoxes, inputSize) {
-
-  const numCells = getNumCells(inputSize)
-
-  const mask = tf.zeros([numCells, numCells, 25])
-  const buf = mask.buffer()
-
-  groundTruthBoxes.forEach(({ row, col, anchor }) => {
-    const anchorOffset = anchor * 5
-    for (let i = 0; i < 5; i++) {
-      buf.set(1, row, col, anchorOffset + i)
-    }
-  })
-
-   return mask
-}
-
-function getCoordAndScoreMasks(inputSize) {
-
-  const numCells = getNumCells(inputSize)
-
-  const coordMask = tf.zeros([numCells, numCells, 25])
-  const scoreMask = tf.zeros([numCells, numCells, 25])
-  const coordBuf = coordMask.buffer()
-  const scoreBuf = scoreMask.buffer()
-
-  for (let row = 0; row < numCells; row++) {
-    for (let col = 0; col < numCells; col++) {
-      for (let anchor = 0; anchor < 5; anchor++) {
-        const anchorOffset = 5 * anchor
-        for (let i = 0; i < 4; i++) {
-          coordBuf.set(1, row, col, anchorOffset + i)
-        }
-        scoreBuf.set(1, row, col, anchorOffset + 4)
-      }
-    }
-  }
-
-  return { coordMask, scoreMask }
-}
-
-function computeBoxAdjustments(groundTruthBoxes, reshapedImgDims) {
-
-  const inputSize = Math.max(reshapedImgDims.width, reshapedImgDims.height)
-  const numCells = getNumCells(inputSize)
-
-  const adjustments = tf.zeros([numCells, numCells, 25])
-  const buf = adjustments.buffer()
-
-  groundTruthBoxes.forEach(({ row, col, anchor, box }) => {
-    const { left, top, right, bottom, width, height } = box.rescale(reshapedImgDims)
-
-    const centerX = (left + right) / 2
-    const centerY = (top + bottom) / 2
-
-    //const dCenterX = centerX - (col * CELL_SIZE + (CELL_SIZE / 2))
-    //const dCenterY = centerY - (row * CELL_SIZE + (CELL_SIZE / 2))
-    const dCenterX = centerX - (col * CELL_SIZE)
-    const dCenterY = centerY - (row * CELL_SIZE)
-
-    const dx = inverseSigmoid(dCenterX / CELL_SIZE)
-    const dy = inverseSigmoid(dCenterY / CELL_SIZE)
-    //const dx = dCenterX / CELL_SIZE
-    //const dy = dCenterY / CELL_SIZE
-    const dw = Math.log((width / CELL_SIZE) / getAnchors()[anchor].x)
-    const dh = Math.log((height / CELL_SIZE) / getAnchors()[anchor].y)
-
-    const anchorOffset = anchor * 5
-    buf.set(dx, row, col, anchorOffset + 0)
-    buf.set(dy, row, col, anchorOffset + 1)
-    buf.set(dw, row, col, anchorOffset + 2)
-    buf.set(dh, row, col, anchorOffset + 3)
-  })
-
-  return adjustments
-}
-
-function computeIous(predBoxes, groundTruthBoxes, reshapedImgDims) {
-
-  const inputSize = Math.max(reshapedImgDims.width, reshapedImgDims.height)
-  const numCells = getNumCells(inputSize)
-
-  const isSameAnchor = p1 => p2 =>
-    p1.row === p2.row
-      && p1.col === p2.col
-      && p1.anchor === p2.anchor
-
-  const ious = tf.zeros([numCells, numCells, 25])
-  const buf = ious.buffer()
-
-  groundTruthBoxes.forEach(({ row, col, anchor, box }) => {
-    const predBox = predBoxes.find(isSameAnchor({ row, col, anchor }))
-
-    if (!predBox) {
-      console.log(groundTruthBoxes)
-      console.log(predBoxes)
-      throw new Error(`no output box found for: row ${row}, col ${col}, anchor ${anchor}`)
-    }
-
-    const iou = faceapi.iou(
-      box.rescale(reshapedImgDims),
-      predBox.box.rescale(reshapedImgDims)
-    )
-
-    if (window.debug) {
-      console.log('ground thruth box:', box.rescale(reshapedImgDims).toRect())
-      console.log('predicted box:', predBox.box.rescale(reshapedImgDims).toRect())
-      console.log('predicted score:', predBox.score)
-      console.log('iou:', iou)
-    }
-
-    const anchorOffset = anchor * 5
-    buf.set(iou, row, col, anchorOffset + 4)
-  })
-
-  return ious
-}
-
-window.computeNoObjectLoss = function(outTensor, mask) {
-  return tf.tidy(() => {
-    const lossTensor = tf.sigmoid(outTensor)
-    return squaredSumOverMask(lossTensor, mask)
-  })
-}
-
-function computeObjectLoss(outTensor, groundTruthBoxes, reshapedImgDims, paddings, mask) {
-  return tf.tidy(() => {
-    const predBoxes = window.net.postProcess(
-      outTensor,
-      { paddings }
-    )
-
-    const ious = computeIous(
-      predBoxes,
-      groundTruthBoxes,
-      reshapedImgDims
-    )
-
-    const lossTensor = tf.sub(ious, tf.sigmoid(outTensor))
-
-    return squaredSumOverMask(lossTensor, mask)
-  })
-}
-
-function computeCoordLoss(groundTruthBoxes, outTensor, reshapedImgDims, mask, paddings) {
-  return tf.tidy(() => {
-    const boxAdjustments = computeBoxAdjustments(
-      groundTruthBoxes,
-      reshapedImgDims
-    )
-
-    if (window.debug) {
-      const indToPos = []
-      const numCells = outTensor.shape[1]
-      for (let row = 0; row < numCells; row++) {
-        for (let col = 0; col < numCells; col++) {
-          for (let anchor = 0; anchor < 25; anchor++) {
-            indToPos.push({ row, col, anchor: parseInt(anchor / 5) })
-          }
-        }
-      }
-
-      const indices = Array.from(mask.dataSync()).map((val, ind) => ({ val, ind })).filter(v => v.val !== 0).map(v => v.ind)
-      const gt = Array.from(boxAdjustments.dataSync())
-      const out = Array.from(outTensor.dataSync())
-
-      const comp = indices.map(i => (
-        {
-          pos: indToPos[i],
-          gt: gt[i],
-          out: out[i]
-        }
-      ))
-      console.log(comp.map(c => `gt: ${c.gt}, out: ${c.out}`))
-
-      const getBbox = (which) => {
-        const { row, col, anchor } = comp[0].pos
-
-        const ctX = ((col + faceapi.sigmoid(comp[0][which])) / numCells) * paddings.x
-        const ctY = ((row + faceapi.sigmoid(comp[1][which])) / numCells) * paddings.y
-        const width = ((Math.exp(comp[2][which]) * getAnchors()[anchor].x) / numCells) * paddings.x
-        const height = ((Math.exp(comp[3][which]) * getAnchors()[anchor].y) / numCells) * paddings.y
-
-        const x = (ctX - (width / 2))
-        const y = (ctY - (height / 2))
-
-        return new faceapi.BoundingBox(x, y, x + width, y + height)
-      }
-
-      const outRect = getBbox('out').rescale(reshapedImgDims).toRect()
-      const gtRect = getBbox('gt').rescale(reshapedImgDims).toRect()
-      console.log('out', outRect)
-      console.log('gtRect', gtRect)
-    }
-
-
-    const lossTensor = tf.sub(boxAdjustments, outTensor)
-
-    return squaredSumOverMask(lossTensor, mask)
-  })
-}
-
-function computeLoss(outTensor, groundTruth, reshapedImgDims, paddings) {
-
-  const inputSize = Math.max(reshapedImgDims.width, reshapedImgDims.height)
-
-  if (!inputSize) {
-    throw new Error(`invalid inputSize: ${inputSize}`)
-  }
-
-  let groundTruthBoxes = assignBoxesToAnchors(
-    groundTruth
-      .map(({ x, y, width, height }) => new faceapi.Rect(x, y, width, height))
-      .map(rect => rect.toBoundingBox()),
-    reshapedImgDims
-  )
-
-  const groundTruthMask = getGroundTruthMask(groundTruthBoxes, inputSize)
-  const { coordMask, scoreMask } = getCoordAndScoreMasks(inputSize)
-
-  const noObjectLossMask = tf.tidy(() => tf.mul(scoreMask, tf.sub(tf.scalar(1), groundTruthMask)))
-  const objectLossMask = tf.tidy(() => tf.mul(scoreMask, groundTruthMask))
-  const coordLossMask = tf.tidy(() => tf.mul(coordMask, groundTruthMask))
-
-  const noObjectLoss = tf.tidy(() =>
-    tf.mul(
-      tf.scalar(noObjectScale),
-      computeNoObjectLoss(outTensor, noObjectLossMask)
-    )
-  )
-  const objectLoss = tf.tidy(() =>
-    tf.mul(
-      tf.scalar(objectScale),
-      computeObjectLoss(outTensor, groundTruthBoxes, reshapedImgDims, paddings, objectLossMask)
-    )
-  )
-
-  const coordLoss = tf.tidy(() =>
-    tf.mul(
-      tf.scalar(coordScale),
-      computeCoordLoss(groundTruthBoxes, outTensor, reshapedImgDims, coordLossMask, paddings)
-    )
-  )
-
-  const totalLoss = tf.tidy(() => noObjectLoss.add(objectLoss).add(coordLoss))
-
-  return {
-    noObjectLoss,
-    objectLoss,
-    coordLoss,
-    totalLoss
-  }
-}
\ No newline at end of file

tools/train/tinyYolov2/loss.test.ts

-import * as _tf from '@tensorflow/tfjs-core';
-
-const faceapi = require('../../../dist/face-api.js')
-const tf: typeof _tf = faceapi.tf
-require('./loss')
-
-window['faceapi'] = faceapi
-window['tf'] = tf
-
-const anchors = [
-  new faceapi.Point(1.603231, 2.094468),
-  new faceapi.Point(6.041143, 7.080126),
-  new faceapi.Point(2.882459, 3.518061),
-  new faceapi.Point(4.266906, 5.178857),
-  new faceapi.Point(9.041765, 10.66308)
-]
-
-window['net'] = {
-  getAnchors() {
-    return anchors
-  }
-}
-
-describe('loss', () => {
-
-  describe('computeNoObjectLoss', () => {
-
-    const computeNoObjectLoss = window['computeNoObjectLoss']
-
-    it('should only compute loss over scores, 1x1 grid', () => tf.tidy(() => {
-      const outTensor = tf.zeros([1, 1, 1, 25])
-      const loss = tf.sum(computeNoObjectLoss(outTensor)).dataSync()[0]
-
-      expect(loss).toEqual(0.5 * 0.5 * 5)
-    }))
-
-    it('should only compute loss over scores, 13x13 grid', () => tf.tidy(() => {
-      const outTensor = tf.zeros([1, 13, 13, 25])
-      const loss = tf.sum(computeNoObjectLoss(outTensor)).dataSync()[0]
-
-      expect(loss).toEqual(0.5 * 0.5 * 5 * 13 * 13)
-    }))
-
-    it('should only compute loss over scores, 13x13 grid, batchSize: 10', () => tf.tidy(() => {
-      const outTensor = tf.zeros([10, 13, 13, 25])
-      const loss = tf.sum(computeNoObjectLoss(outTensor)).dataSync()[0]
-
-      expect(loss).toEqual(0.5 * 0.5 * 5 * 13 * 13 * 10)
-    }))
-
-  })
-})
\ No newline at end of file

tools/train/tinyYolov2/multibox.json

-["6ff8b403d437ec9210aed005f4b7d5de.jpeg.json","701a7573675975ffa17b5958d994da04.jpeg.json","70305901c9438c83241ded45a614a05a.jpeg.json","703dbe721760333b14f219b071af0258.jpeg.json","70644f2d900f3e3df8a094514378c8dc.jpeg.json","707c483fbfa6e10ea272052dbac91fbf.jpeg.json","70a226ff7b913d022f432a6262c432d5.jpeg.json","7103f3c81d7b9b5f47607f5517aa3210.jpeg.json","71040961f9178cb641ee2bda77daa1f1.jpeg.json","7153af94856dc09d65ca940b407ae437.jpeg.json","715f772a6a2306fd040598065dbe80e8.jpeg.json","71ec70fdebfe01fc82c055135e73e5aa.jpeg.json","71f38f7c6fd203dd193725a6575eb84f.jpeg.json","72058ea0b2f0ea32f6c1534dd0f4259a.jpeg.json","73203f4c115da50ef1bc411a2a043d22.jpeg.json","748311686961c891f4afbd8a373c02e6.jpeg.json","7531f8ba7c212c6ac13bdb1d6fc2a19f.jpeg.json","75349eb64fdd5d0fbf005df5c7594f07.jpeg.json","753681fbf3d3b2808b8c889f215feb5c.jpeg.json","755cb42ce7f1df46a1b893ea5d26cf26.jpeg.json","75f27276f628e427a2bf80aa7f9792e3.jpeg.json","75f2ace4925a2a749d6942885766f036.jpeg.json","76016b537ef69992905dad5c40289e21.jpeg.json","7624165ba78d723987fd7734ba6af201.png.json","76490fbac3548a57ce43fed12b29d5c6.jpeg.json","770ce887b2dc0c854d100ce8f11d30f7.png.json","77e3e59d693d7071abc68c4428802bbf.jpeg.json","7803ba050aa011496bb49f45ae395617.jpeg.json","787fd9830b9534d4603c704921312dc5.jpeg.json","78aa83fb0a9cd9148258af5201be89ec.jpeg.json","78b8ccc845900d7795b47d08618a0cd6.jpeg.json","78d5b85760cc495385c47f9ae486b37a.jpeg.json","78f3dad6dd4ace4ce16c1520c0c79a1a.jpeg.json","7941a02b69f84c5fe9faaa94510b2860.jpeg.json","798fd92d7074c29049fc772bdcb371ab.jpeg.json","79a1d77dd1380dbd15299ed72dfde5de.jpeg.json","79b0035db4d98d20736f354de4b7afc9.jpeg.json","79d6ffadb928a9aef20f1d3cd68810d6.jpeg.json","79f2078ec3fa4131062ec733c8d3ca10.jpeg.json","7a364f7e14fe9106787773b4bcb5cfc0.jpeg.json","7a6b1f70f5bd8e18f18c0c17a92848d3.jpeg.json","7b68ecb287ac3fdc8af59ed8d543848f.png.json","7b741584c0f339e91ecb7413c306d084.jpeg.json","7b7996d595b2df7ef2fd1a5502bb83cc.jpg.json","7b90dc94a75f26bd402735b26f45d017.jpeg.json","7bdc2b31b0632aff9922497b144972b3.png.json","7bdf2d802946995231b14274b2405ba7.jpeg.json","7c4c6e1c4cd562c7fc692fb3740dd1fb.jpg.json","7c5ce28f52c8befd9f86e6781a02b8c0.png.json","7cca72949282cc482fbb57b6230a92a7.png.json","7d3b0acd6672f0e36bd70153e53a07a8.jpeg.json","7d3d5a022ad59550f3ef259236e826bd.jpeg.json","7d514ff53e1150d4c1744aac4a2009ff.jpeg.json","7d52f80060a528bb7b2544c04baa65fb.jpeg.json","7d78e2881e3a2f454ce2cf822ca3e6af.jpeg.json","7e0ddfcc80d573770cd39f5853ec8daa.jpeg.json","7e27f9f4225299a55f786ab6d8eed544.jpeg.json","7e6296f2e01c6211eb7c5ba74ac3f4f7.jpeg.json","7e68577d4a092b66b009fe16f89b93ee.jpeg.json","7e9247a43ee9469014d72cfb7c6d23b9.jpeg.json","7eb014c79aa03c18f11a58ea58612c19.jpeg.json","7f9b22a160f257d7b4d3a80c09bb38ac.jpeg.json","806a1e4b1ed5a58f6b541fddad7f6a21.jpeg.json","8071807fba2c90fd85714cb40048c71f.jpeg.json","80787772d05ff9f8e58fc3e276df2c8f.jpeg.json","80a32fbe55da9c650fad4211badb09dc.jpeg.json","80ec0adfb26fc2d9e702382685560682.jpeg.json","8144caa0d00ba83634057866be21bb43.jpeg.json","818da784c33a36dade6ddbb4acb62a96.jpeg.json","81b0e7e46da0e150b295ee7a6d3e1d4f.jpeg.json","81e4ba2ee066ff09ce661cd1d9750ef2.jpeg.json","8202cf2736f818a34875d753d326b2e9.jpeg.json","820e93571774600596cbf7d152c12bb5.jpeg.json","828dbd12d2ed9d0ab59c39e0cf52405d.jpeg.json","82d4d7b29cdf9d02883002309f0a2b51.jpeg.json","831b1a75dbf1f605ac934cac47f2f6d8.jpeg.json","834bf7852cb1b14d1e1a3ba2a218bcda.jpeg.json","83631e520695495b92607c43168db581.jpeg.json","83c35f6fdf42ab14105836ede211a16c.jpeg.json","83e41860c05866c0e3918652417f8e45.jpeg.json","83fa5b54a47101cf0f63190eb24e1cd1.jpeg.json","841e1d2566656bded15d5e88c57c030b.jpeg.json","8441ce1f48fb9f42c1dc1965cc8c321f.JPEG.json","8475b7301e3b286871edd56452fd8e3f.jpeg.json","84914d767fffb2ba4fbeb2e04bf59d0b.jpeg.json","8494de96ec43cec4d3d2c33db672f403.jpeg.json","849ba4e71a31608b8dfa3407f50bf9a5.jpeg.json","854b6eb006aedb7c6ce72bdf5a078d51.jpeg.json","85838401549e4a1f22555627f44cc955.jpeg.json","859df8e5277a1c70815df5a5ca9ecffb.jpeg.json","85f59c64d02330962fbcfa26d7abfef7.jpeg.json","86040943da555695cf999b4ab9d03bb0.jpeg.json","866f7191af186f97a1b8c9623d73a82b.jpeg.json","86854f0064b17b918f90bc57a1980437.jpeg.json","87b431be307a704329f32ee4a63ff10d.jpeg.json","87ca351b10e672fbb1b3dde77e5525e4.jpeg.json","87df1901f592db9f071adb357b3c2d42.jpeg.json","884326038ea0d8ab9a9aa7847a87b73b.jpeg.json","886ad98d2d123c4192efdc0a00df4635.jpeg.json","888a1119743c85ff9120870f1c10fa53.jpeg.json","88badd516a416a5dd1528759d81dfa2f.jpeg.json","892e21ec3aa8c1ed1765a431b2b68482.jpeg.json","8976d9fc2c5b2ababda1406d043bdb74.jpeg.json","8a4e32e64a7dd08e29804623d4b96547.jpeg.json","8a59d79dae544c9046fd12f476c97165.jpeg.json","8b4bd50543e08290446d32a2a12f6ffe.jpeg.json","8b56fb59e1e1fd22a2df2013dd438c35.jpeg.json","8ba2afe0a859ca8156b8310badca4c9f.jpeg.json","8bb8ab40226c6368f143bf29c1bbd5a6.jpeg.json","8bdd8a6919661ea6cc729a31ab0955d5.png.json","8c3b0c73ce253840febcd51f3b8e8ce6.jpeg.json","8c5f38f5ab076909e00b447248d00687.jpeg.json","8c9496dc6cef4a0983fb3973b55b02a3.jpeg.json","8cb75fbf0624fd5cd086a49b1be8b2bf.jpeg.json","8cbe52a94b0701078098008e8f95c6ed.jpeg.json","8ce2b5a1afefdcbdd3462813a55090df.jpeg.json","8d0a85407a1eb4ed4162030d482275f9.jpeg.json","8da7129adbc4cb11a0387d3542fcf615.jpeg.json","8db1f21462137313d55f1a162ae1dd76.jpeg.json","8ddba860f36f70d82496d0f9d0b88bf1.jpeg.json","8df2cacbafde9eb289e229ea80f156f9.jpeg.json","8e2a21e0bf7fcb73adebe3dc760aac3b.jpeg.json","8ec4f340ff26d4d3685d55a9e51664a2.jpeg.json","8ee0a6ee1d87e42c377ff2b96c282f02.jpeg.json","8ee56a12c1f8929e0af9ed466e798381.jpeg.json","8f06cd4ff26dec9e56ccd07220446075.jpeg.json","8f0e47302b522ed20a7520c5b4a78f7c.jpeg.json","8f618dfa969081acd3012b296e2201bb.jpeg.json","8f9b1e321b35dba0d6942180979cb66d.jpg.json","8fb501f2238f9eb0d2d1026758cd15fe.jpeg.json","901c453cd81a877ace73eb0ad746a2ed.jpeg.json","9032a049b9745b29b5435ed253a49ff5.jpeg.json","90a2d791c096a54f78f0b8a04bbe8da2.jpeg.json","90a7881307bd13eb4bc9be47aeb20f28.jpeg.json","90c4545574be9de286262b7eff20adac.jpeg.json","911f4132ffbd482d9eb7f85f66b2ecee.jpeg.json","912f85ed5abd31fb7adb24c2bd8859d6.jpeg.json","91b1be0ae205888ac0b77ea6320e2ca1.jpeg.json","91d9f5ab78276a70f0972e3d8a15f745.jpeg.json","91db4d6a662b759a72d48707ba43e34b.jpeg.json","91ea2f5359cc19280d96c101bfe22759.jpeg.json","921a572517b61ec2d8096256b22fbeb2.jpeg.json","9236e75942f85a5a014e93aac73dfeff.jpeg.json","927810f080a7df039fb429fc1034c146.jpeg.json","92d8769ab1ac7919e46086e267349ece.jpeg.json","93f5fa9beb051efc61b931463b3362a6.jpeg.json","941afb668b7b5432667fc0cccfbfb641.png.json","941c9453c25b97cf56b81daf0fa3e2f1.jpeg.json","943cb181858d995be5225d1419e50c5c.jpeg.json","9441715282a08469d681a97ccd4b233f.png.json","947ddefa87430eaf0c40729d0fc7410f.jpeg.json","95355a69b7b39c60259719705cea9900.jpeg.json","9558b20aaeb1b1c833e20ccdf413fbf7.jpeg.json","957cad737d5e939581f3e8aa027233e1.jpeg.json","95c563f18662d1987996cdb2ead4624d.jpeg.json","95f5e3159aaebf5831ef0409cf274075.jpeg.json","9639be2d1eff02ec7162a823e669c44f.png.json","963aef3fff65546cb9803cfaf889ca34.jpeg.json","96440c01c37a3a594dd42ef21431f3a9.jpeg.json","96450269ec8e68b0108e8b5b4d145c55.jpeg.json","965c8a1ffbe3fc9951ee1ef5e34c6d37.jpeg.json","96ab72a688f57d0de88c71ef0bb6cfe7.jpeg.json","97738f79dad5424d6e1a440a3a461532.jpeg.json","97b8f38e9cf3adbf6657b25ce1a7335c.jpeg.json","9813582798fab064746a1b94a0f38f24.jpeg.json","98ccbfbe45bc1bcfbc3f40ec39a5f41f.jpeg.json","98f688756e56414609e0e27d39701c56.jpeg.json","99366e7ee9249b7dff69856eea8bd9a1.jpeg.json","993834e21792dfc43bfa9dd6bac01013.jpeg.json","997c7bd5d32ec45ea86a13b11444a59e.jpeg.json","99e22017232b6de400542df046a784ce.jpeg.json","99efc5d7227e94e8fa5db14657440d93.jpeg.json","9a41ce7c515fca5f0e3d852553bf7401.jpeg.json","9a586f9a33601864bd887e6a3fc15306.jpeg.json","9a9adfe19d6c60cf14406ee7cc4ba8b9.jpeg.json","9af6d1a0b444caef74043671fecc5a2d.jpeg.json","9b01e5e1794f2049e77f62be7d61fefb.jpeg.json","9b09ec48138fcc10bb6546e561550205.jpeg.json","9b71abcaaf16fb9544330871c1997c61.jpeg.json","9b72b7e9c7a2eba1225a2b94bf6c075a.jpeg.json","9b74584ba01059866029f10e04a28d86.jpeg.json","9b9f0306056eba83ee63fe13309d6416.jpeg.json","9bc01accf2bd998927e35eac49fbe701.jpeg.json","9bc603c6f4135cd85aadfb0ce280cf2e.jpeg.json","9c11c8da895be5e51f6f376bd565a19f.jpeg.json","9c1fa32f357d296e353e8eaa1c0de5d5.jpeg.json","9c417c45c7c2f840657b66c5c6445e1f.jpeg.json","9cfe50b051a74f0d34d47b2c2fc289ff.jpeg.json","9d833f506e39a9b256b7895ca4f4a12e.jpeg.json","9e6788e49029872d6af65628e47dfb42.jpeg.json","9e753701f51917034425ec42089227d1.jpeg.json","9e94808c0f99648e03ca95de2db13aaf.jpeg.json","9ed3aeff287805318f9158e973a84d80.jpeg.json","9eec8bf302a0b554a97ec051326e207d.jpeg.json","9f0354346e5f10f3a9cfa52298f988af.jpeg.json","9f470ac5164bde63ef3be45e9710e55f.jpeg.json","9f814896cba97063fe41f14519f3e3a9.jpeg.json","9fa8ba16bdb2e9866dc88edb3d6d4dbe.jpeg.json","9fb879fc041e93f09f4d4220f54ac496.jpeg.json","9fc6ace3668c28d8843ce3faeb967a81.jpeg.json","a11192978ca202719cd557e43ed49985.jpeg.json","a2985d563eeb38e0cd9853d9874d2186.jpeg.json","a2a3ec05647e4708b65b988b72831a99.jpeg.json","a2c345b97444160ca424f69ab2967424.jpeg.json","a2e6f5eb92c42c6c081375193e4016c0.jpeg.json","a3078d34dedd5f24001b5e0c629e53d1.jpeg.json","a36349468a0f3303e247c3bf1a3624f9.jpeg.json","a36cc42841175f14579ff80d9c21c7f8.jpeg.json","a375ca2f02bce027332edc0b59412819.jpeg.json","a43687a677c502eda3825a8e22504865.jpeg.json","a4a36933aa80f535d0d26beef6d534df.jpeg.json","a4fdc6f8ff15a377a7bbef763ce056b4.jpeg.json","a50e6b1510a72ee28a09f8f89201a515.jpeg.json","a51cf5fbcebabdb1b9b108349ef4c6e5.png.json","a56531bd2a7405fa41ec2c89a7740772.jpeg.json","a56e1286a9516e4c2a1ef113ab6b7f9c.jpeg.json","a6284ea25b31c6d9744e30e430ba22bb.jpeg.json","a64bbd93b3c31558166c74267c4ac73e.jpeg.json","a701b6c44062f5392e9375b3a70211a1.jpeg.json","a7582f2944bf8b2dbffb613d5e81911e.png.json","a768268f66f7ef309cb5adce6932e916.jpeg.json","a84850991d22f6dbcdac714225d2ed5d.jpeg.json","a9168a323a9861c5d44a27ca12cdcba3.jpeg.json","a99c7bbde59c831006c02ba047b3895b.png.json","a9a3c66d69397aabc475dea5b3d46fd7.jpeg.json","a9c70118fa87e20076152f8df3303e0b.jpeg.json","a9d60d9edf124b2a6e0371a9d4b1105a.jpeg.json","aa21c7f303b2b5e161883704901ca315.jpeg.json","aa248053b9f210e59cebe829e3c9b7d5.jpeg.json","aad7d6917fb9297e3647719963062f23.jpeg.json","ab10e5cb60fe670021c78c78adbab5df.jpeg.json","ab613048422816fdf3c15608065a29a1.jpeg.json","aba73d12fbe9f8e1b5477a197e07dfb3.jpeg.json","abdd45bcbb22860d36069b0019247198.jpeg.json","ac1743e4e6991795d887cdf34e7f68da.jpeg.json","ac294fcb10136cc069ca76fea18b1012.jpeg.json","acd9f38ee4702a6539408c69f593e091.png.json","acee560d56cf1e1c3be2fe5cbce85311.jpeg.json","acf27a1a77b1c64826d23fdd8b314de4.jpeg.json","ad8afd075c3ba5a3ffccb866223f5edb.jpeg.json","adcfa03f6ca674ae5c597dc120e76a89.jpeg.json","ae35ae8e2be13b5b45595c274b1cf7eb.jpeg.json","ae452077f9971d9921a7afb3922048cd.jpeg.json","ae6df2a0f14877bc5461abc3340e50c8.jpeg.json","ae6fc6f56c8727760e79d8c2b68036f2.jpeg.json","ae6fe194927e756df7fe631522f1c610.jpeg.json","aef1eeb4fd79ccf182859a3246c1d8d6.jpeg.json","af201f933bac670b2dc2350e86cee5c1.jpeg.json","af3f7fde9d22f417cdc0df4753e1cb50.jpeg.json","af7a777c97af1ad15984c4a5301ebf33.jpeg.json","afe3e53779e69ca42aab7f1791a5da39.jpeg.json","afe5047af1f74252f2c9c60992b91056.jpeg.json","b009a59ff072a43ea399a6e5ec517abf.jpeg.json","b099b4fb9dbe29bfa7c7b6be1940b614.jpeg.json","b0d92bfe4af109347d4463c46239d39e.jpeg.json","b1507b84f010a6a9422dfc64fb543487.jpeg.json","b16ed70eed5fb43186894fe692e7d475.jpeg.json","b29f34f5ae0a0809fcfa977b5ad9da87.jpeg.json","b2b7b90e86c5825806925d455e5bb216.jpeg.json","b34a507a4e4cfb5ebeb22839e3720d9f.jpeg.json","b369335407124257410680bc75a29766.jpeg.json","b3732c175eb8b6f9cd1ae9045166e16f.jpeg.json","b3e5dfed31c4e7bc8effbacf61ff82c8.jpeg.json","b40755769b8c56c01dd68b0dea7845cc.jpeg.json","b42335a9da636e4ad3cc00d3cc4ba05d.jpeg.json","b426df0cc4acc34ec5af159e8d42efd0.jpeg.json","b43941538c97ed7b14e4b5247f791e41.jpeg.json","b449be070cb8539436be650255bda41f.jpeg.json","b4a8395a54e094ee1250860e34c2ec07.jpeg.json","b685e3383780ffeacb8e6297cf0ae4b3.jpeg.json","b685e5f2be5e93d34e8041c66c1f1e03.jpeg.json","b6d011b0bb1a62abaab18b468449f784.jpeg.json","b6d427bdcb4ccd66b43f2bcfeea2192b.jpeg.json","b6ee014f5d623c6cca88732ac6d41c7c.jpeg.json","b71b3c9a493a1f63adb497cbf6cbfb2a.jpeg.json","b7355aa0b18badfff95d5fd5af6b8268.jpeg.json","b7616310f7f218dd57906e6e0cfcafa5.jpeg.json","b77a8535c138319408a0c757fffb1d79.jpeg.json","b8688ea021519777f61f124a419fdb29.jpeg.json","b86b50c9f0ce0b64431c4081a2f573dd.jpeg.json","b8ae48811862229e89f88794a2f02ddc.jpeg.json","b8cf36cfc7f9ae5a48714e475b25f164.jpeg.json","b945150c2ec585f3d7bd0a9102f8540d.jpeg.json","b95079e842b08ff260630efe2fe7e141.jpeg.json","b99f523d94bb7df35b7d028fdf94e988.jpeg.json","b9ccf86f5471af4d6428f873d8d958a8.jpeg.json","b9dc2bea6f9241ebeb5fc7e7ff8c04c6.png.json","ba1071f184a84439df9f6114b6e70472.jpeg.json","bae4b514eed98090d7e9aa81e0b03de2.jpeg.json","bb41fb0ae0d6f9a036a56def9924f172.jpeg.json","bb8648702ffaba9723ce164f24ca16d2.jpeg.json","bb8e78ad2c05150f55ce85704c4023e5.jpeg.json","bb8ec58ee700b811a205f577450160f1.jpeg.json","bbbef3b7c7943694c0969e11c35179ad.jpeg.json","bc390abe92c933d67427607a90c5d8e0.jpeg.json","bc3c5a9d4543ed7f5795582a632b20ce.jpeg.json","bc437be6d3ee896e081a752a7f7dd9f1.jpeg.json","bc6a12c985b8ee2d74d4c9bfb9918791.jpeg.json","bcee2f70904deff9fbfa47186c382dab.jpeg.json","bd78ced66d4e9939f7c4dca321c5cc64.jpeg.json","bd88325d1e6c52f87981f7b88d9f4efe.jpeg.json","bd92ceabd72484a556792c0efefc9071.jpeg.json","bde5e4571fee9681d6e7f2e089b5e4c5.jpeg.json","be26e439c5348cbcf8802fb66ea10ade.jpeg.json","be5c23545eb93575e7060f01f7ddcf88.jpeg.json","bf98d5ffa7eece6b3e72683e29c184f7.jpeg.json","bff24593f4c61fd4cfe25a63f019822c.jpeg.json","bffbb92b56845b871a727abb6e0cc9b7.jpeg.json","c03a255bb791b438134bcc211bcfc229.jpeg.json","c064de930c0c79937e528b0142a30da6.JPEG.json","c097e31ab213847094af00c163c27223.jpeg.json","c0bc62c518d17b377e60b8c6c046b862.jpeg.json","c1505ff920b8712d5d221e9ffe70709b.jpeg.json","c17e47a687117911eaaab9eca760c4eb.jpeg.json","c17fab603ec0c74155b1636c4418696d.jpeg.json","c1b15b94b3dfb7c964cc0027c8258ee7.jpeg.json","c1eb726692d48e56505f69d1d5e8f93e.jpeg.json","c238aea63cbddfbd121c944efc62af6b.jpeg.json","c25288d3b882fc081156387819203347.jpeg.json","c2e97bd83465c1013fb449aae9c139ba.jpeg.json","c311ca4d933f354d9a1b7e1a0ce5a158.jpeg.json","c37fcd546f12a1027b3aaa98ca49b567.jpeg.json","c4709b17ce7ae094ef8d148567aafc12.jpeg.json","c4dd7333fd768946314126dbaa209044.jpeg.json","c5428858d364875f454de16ea20209ef.jpeg.json","c54ae5db16c59a7689c3181b794776c5.jpeg.json","c57c739aa8fdffc25b3df14cb5e8cdf2.jpeg.json","c5c9c9f3078b29003c3eed19a9043419.jpeg.json","c6aca7bf440fd6f83df844419739ff34.jpeg.json","c6bf4ff2724a16eba0c61405e8739b03.jpeg.json","c6f987545f6ac44a7fae4b0d90c1cc22.jpeg.json","c763aefb9189781ee2fb49768f06e2e8.jpeg.json","c7988e0dbf6876b324824d8d58103fba.jpeg.json","c7a5775839e1c03994860dc36bac13f4.jpeg.json","c7ff714283d234f96601bcfd20c3ec03.jpeg.json","c839ef18856b89f21e79652fc322b936.jpeg.json","c862eaddd35a6123ce4eb3af530525b3.jpeg.json","c8702d40118aeae6ae328fe2b90937e0.jpeg.json","c8cc592e53418976b7d906ea53a20bcd.jpeg.json","c92dc36627b6ef845c0430b314baa1b5.jpeg.json","c9452a4f57b646744407a8d6f01e5e59.jpeg.json","c94ff10637406f0aa3c904f9913a7fbd.jpeg.json","c957f1c5f4873e6c3654e0e859ca2887.jpeg.json","c9af4e35587383ea25a9bebb1fc7b90e.jpeg.json","c9d7d4a01962409f8660829649a266aa.jpeg.json","c9fc3d14fe098f8a60e83ecf05072f8b.jpeg.json","cae0f91f1011966b709c653763e689cd.jpeg.json","cb579a43c90f5ac0c067523dc3a4d439.jpeg.json","cbaa4ba0460b5b82aafeed1065407d55.jpeg.json","cbbdd74c5e5ae3ee1ca65d78e1aae440.jpeg.json","cc312bc2f7d212f7ac16d680a08d132f.jpeg.json","cc3f325ba4ac7e580a93a7a21df57ddd.png.json","ccdc1da1cca2d2037f4fef94fc043853.jpeg.json","ccdc67e21001c5e546dcac8449598539.jpeg.json","cce9be1ce40f81e773a4e881ffc86470.jpeg.json","ccf563a0e70c62e3208854ef7c5b827f.jpeg.json","cd2d43a111837b154397252ae669f3fb.jpeg.json","ce382ad55ec548e681cf4a04a59a5580.jpeg.json","ce3863ca96e401fbba5adf438d98756f.jpeg.json","ce5b568520be24a630d7d6a7f9e4bf80.jpeg.json","ce65a1854e71ea6c4bb7416088713ebd.png.json","ce7318afda1c720cfa8808d4a4b512f0.jpeg.json","ce8110205ea28a8e18aaea50ac566fc7.jpeg.json","ce8519acff3b6c7877ffd964384ba7bc.jpeg.json","d03f62e47e2cc64f827d8dc3c9a56882.jpeg.json","d08e0703342cadb0cb4aa2251afd601b.jpeg.json","d0b07855a47678b6a8d7ce4552771c7c.jpeg.json","d0c5622bc33161aa3721cd7ded296c9d.jpeg.json","d0f3c17be64e044984b83b9e4aee4e3e.jpeg.json","d114f36c25a348f912acc7aed1f05d23.jpeg.json","d133c00b5989c9b453dc30b78cd9552d.jpeg.json","d148850ed7307574db17f9d965a2bb90.jpeg.json","d16a63cf6ba1a70d910a4c6689952a7e.jpeg.json","d18f1718fdc29baf281eac12b3a18b3e.jpeg.json","d1b4fdd0fe09e829944f8dad6efb0f4c.png.json","d1b72e014226de6f9b9f344e02860830.jpeg.json","d2130727129386cab1ed53a87a56e239.jpeg.json","d2e4e8808a62c404f38498faf3b86abb.jpeg.json","d310d9ddb154fbeeece93e5e85940bbd.jpeg.json","d334da1da9ff6e2975664b8005e265db.jpeg.json","d3ad82ac8936c1a68a3da54f7bd4217b.jpeg.json","d3c482aa1c424068f5e4244022b0ee49.jpeg.json","d3edfe6bab836691ec79f363a5c796f3.jpeg.json","d3f7d6da6e9ef8fd6dbbeff7b05afeaa.jpeg.json","d401f57bc3aff32e363b3168dfa86e62.jpeg.json","d44d401b12427837b14c734be51e0aee.jpeg.json","d467ab7b3d8beff8d51024d3f6ccf6c3.jpeg.json","d49ec827d0b340989f64be6fc6f996f4.jpeg.json","d4bc6f134ef7d2e087be9cf908c2a2e6.jpeg.json","d576adbe497d0859636e9362565cccb0.jpeg.json","d585bcb7110dbbd3d2c448fbd56e1de4.jpeg.json","d5af05d60ec4ecc0bfb110f58b175ab0.jpeg.json","d6330df9b0f62cc9a66b5bfc8bbc9949.jpeg.json","d657eb1e8aadf3c89cba389a6c118ab3.jpeg.json","d6913dd3374d981053042e99699aecde.jpeg.json","d69600cc178936838fb7fdb1118e7537.jpeg.json","d6a7b438e7dae8b357012f006a5853b9.jpeg.json","d78df2ab61897b99f3a08831ec348e45.jpeg.json","d7aa773a4dffafe161c0e30de3379324.jpeg.json","d802cfd3bbbc3276258ba689c6999a32.jpeg.json","d81e16b7e522222cd27a51b483c11800.jpeg.json","d87a53001c3cc399b76a638f731d2cc7.jpeg.json","d8ed135ab9d50e136394a2256a6c9d46.jpeg.json","d8f8f0507a48b76ef59060bfa9b2cf47.jpeg.json","d915c4b42c400f9bbb7ed5a713375e86.jpeg.json","d931f074be18d83bc76fad43dc165081.jpeg.json","d9573cbf00349536974d9ded8223e4c9.jpeg.json","d9c2b4df4f06e4715feff4f50a073a03.jpeg.json","d9e725406bee40d5acd6c8042031855e.jpeg.json","da14aaf7d6c01bc230a5d2c82d1c9500.jpeg.json","da233e2bad57b401c494a4f62883e34a.jpeg.json","da896edbf57d593dd39247b5a5965ca5.jpeg.json","daa446ba29c460532f352814f4026685.jpeg.json","db200ab7ac7cf86e7cdf86da6347edc5.jpeg.json","db3266e571b66e3ce100c0033710cf1e.jpeg.json","db61ad9e74d0f52204608b894dccd498.jpeg.json","db656485be6ccaa7163c9561ce8d4db0.png.json","dbbee0ed60f3d78f7f3d40c4e9f04b05.jpeg.json","dbcce7ec562b55974af45addcfcdb291.jpeg.json","dc1e1228d473d40b6e41712a5db17366.jpeg.json","dc2c97305ce4686d135c86fa48c204f0.jpeg.json","dc6f2a9dd6fc900c8b1e17c33eb9e2b1.jpeg.json","dd2a8c3d9757912e71a7b89466b0dbf9.jpeg.json","dd3c8b67685c6a59b876a3e01d393d13.jpeg.json","dd562c58de783991b44578cfcee09ee1.jpeg.json","dd5658d1decea90a9af7a37d1031f8c6.jpeg.json","ddc0ed9c03c92a1d89064132f8be0403.jpeg.json","ddef41bc01129efd1f84bb3555bca104.jpeg.json","dec14706cfee632134eb72fd3821db6b.jpeg.json","df0c46342a07183d8bc8c73fc7a8807c.jpeg.json","df2264b8511a4ceff8b708982e4cf4c7.jpeg.json","df529f96a1a27b98a412aa6507c967fa.jpeg.json","dfb5016b199c5c7a68af6fd54dff8a5a.jpeg.json","e049759552aadbb681faeea21a386ddc.jpeg.json","e065afa7e8354ff5d2fa6b8fc2e7133e.jpeg.json","e08cbdd522c38bc9622d4ac7055f2f13.jpeg.json","e09f9a594a5f082e1b5320dcc200f73a.jpeg.json","e0b13dae2ee6176ea0997a69c73edb6b.jpeg.json","e0edee1feb02e46a737d785731660606.jpeg.json","e1627a23c077a873c8b3feaa0865abe4.png.json","e1819aae68143e22365034525c067dc8.jpeg.json","e225981599523f56c082cef434378250.jpeg.json","e26b4a319348af33c3eccaae584fb273.jpeg.json","e2cbf82cd42642f86a79ad520cc1ff4c.jpeg.json","e33ebc36feb28144cc11e13a3d554625.jpeg.json","e352f13f8d13bb47833807e01f1ed4ed.jpeg.json","e3961ea3be6f5072723fa8a9b481ba35.jpeg.json","e39ec5e9d0761cb2f42b6fd951a8cdb1.jpeg.json","e3a62394af2093967325d9c48fe96e97.jpeg.json","e3e614d08c544a28ecb766e5dd2b0285.jpeg.json","e43ab595eb777ac412a9234a6fb289cb.jpeg.json","e45c0e2626ef36f047157c73dd8e043c.jpeg.json","e45cebee744e770e06250863575a61be.jpeg.json","e4fbf2a865ef90de7ba12b8a20b16ed3.jpeg.json","e506d47988a469e048f0d30c87a77269.jpeg.json","e51fef4ae33be241c98139b9e5473b0d.jpeg.json","e5578d64e93286b225f8889e57243537.jpeg.json","e5bedd47dcfbe72338ed8887326dfba0.jpeg.json","e5e536c80dd4645521d4b116a323d23b.jpeg.json","e6197415a8e971b2929c42c77eb935f0.jpeg.json","e64a0a52ca7510f358a66c826dd87703.jpeg.json","e66835e2dc2976f5f89c7b431a25b18a.jpg.json","e674947be8a06387107332df62f0f9bd.jpeg.json","e675b5f02a6ef6f054d915aa40b0376f.jpeg.json","e6a2e3f440439eea3d71efae67282fcc.jpeg.json","e6bcf7cf0fa9159a0cda27a29142de91.jpeg.json","e726e0061a9c28d232ad4a82075f9163.jpeg.json","e76165960c6acd2e981542adb051cb88.jpeg.json","e77a77e463d77b7fdd5493ba34e07ae1.jpeg.json","e786698b884afd2b4d2083f54752d13f.png.json","e78dba436cc506b491635587931efa45.jpeg.json","e7a000d9dbc511424ff1f4dba6c0d66e.jpeg.json","e7bdb14f10d4d024113702a7ca8661ee.jpeg.json","e81b86a127345c34837a0fd6faad4bb3.jpeg.json","e82a9debbbd86a16d08c7bb89d1868a7.jpeg.json","e89220fdf4a509c52ae62ec57ebeb8d4.jpeg.json","e9181e87b3fedf0371ab1d22f9e98cd7.jpeg.json","e9266d87658a25bcd0ee1a6000bf7c3c.jpeg.json","e94e1762fbdfee04b1b882cda844175e.jpeg.json","e95cc2a5919596c9411f30a5aa0533f2.jpeg.json","e970e9e2ad1ecffb03ea07bb847d244e.jpeg.json","e9b245358b4b5e1cd1dede8623db322f.jpeg.json","e9cc8329e12b6150a29bb24f7ff7baa7.jpeg.json","e9d8805725a216a4e90bc68abffee7f7.jpg.json","ea1c9a1224b1ebf21bfbcb183041967e.jpeg.json","ea5164f1625fac7fc25785146105026e.jpeg.json","ea5255de62e00538f7a5552d95ffacd2.jpeg.json","ea848887a2d74fe043f81f56b88a2d12.jpeg.json","eaad3399669a44432d88bcefb80dddea.jpeg.json","eab3e34e6ca878e5b16dac1d1e1a6dfc.jpeg.json","eb8c79d310d7054d37890d2290b85bf0.jpeg.json","ebc8c16eb8ebefa11516f6d9b2990edd.jpeg.json","ec23245f92b5a1bff7db98bac02a6cb4.jpeg.json","ece1ca7a319c25f3cfeec23fa97d2fc3.jpeg.json","ed243bb698dc4b19ce281a770034f9a5.jpeg.json","ed4f181c6c8863f2ded095700bc2c7c3.jpeg.json","ed6409e995cff02a1f5a30a024326aef.jpeg.json","edb4ec033232cc5560d21d1a91028ac6.jpeg.json","edb5618bb7c3ce70faa9cea170958efc.jpeg.json","ede4e9ca14dbb92cdaf95576103ac6d1.jpeg.json","ee08415d8550a2d87364de650a07ecd9.jpeg.json","ee6399df5a80d9cc28c7b13cec0abd1f.jpeg.json","ee819fb1e536987c8c566f62791ef8d9.jpeg.json","eea87fd1a1ddc013631567ad9bd670f8.jpeg.json","eecb407083fd3cf5c704f6b0d86ca8c8.png.json","ef3737f8de72aeb3e967b8344be0f5f0.jpeg.json","f0223dc36cfb60e12b176050796453ba.jpeg.json","f0c7529b755b21c1c7f775413203e49a.jpeg.json","f0e8b78f1f18daf76dd4fc9d04211720.jpeg.json","f11090e810ba8d57814a572d78d40b56.jpeg.json","f17b12b2416962a2d668c8899ff2f0b7.jpeg.json","f18e6c5d2c65873cf38884f41489bc0e.jpeg.json","f18f8f3e6d3d18bc98b0818ee43b32d8.jpeg.json","f1d0d1ebe22add8e060e73cf30048f54.jpeg.json","f1f48ffb9b03fdfadb469dd198aba7ff.jpeg.json","f256210bf91062ad03c442bd67e8ae5c.jpeg.json","f264ae3144218ac7e01b5c9f8c8ead0d.jpeg.json","f2cc05d2d2d65f4fd123fc30197d94d3.jpeg.json","f2db964b5fc68cef96db0381a3b3ba77.jpeg.json","f2edefb4489da7af9e3db508a4cdad48.jpeg.json","f338b2e04ed7dd590a786fa7a0466b3c.jpeg.json","f344d8a79230d6eaa5a79d6fe937807e.jpeg.json","f3460f571c21ba0024630a81f7bfd646.jpeg.json","f3c689b454a7ec1c0d4a656e5c73627f.jpeg.json","f4137bf001ab4487db97aa7f8c02f02c.jpeg.json","f485d8be1fa45e835d1180975a32cf0f.jpeg.json","f4d42e49993c2ea60788c5080b8bfd60.jpeg.json","f4f73b8baef2980de373858ae764234b.jpeg.json","f550e853f302025076d54ebd00ce5be9.jpg.json","f5604bfa425414845c84cbdc5f374d01.jpeg.json","f5abe1eac1b1a79fa194e6a6afec45fc.jpeg.json","f5d8576e4c189d11d057a6412dc81171.jpeg.json","f652d7303b131cd7531f1b13e8aca173.jpeg.json","f68d939fda43324776d14209edcee17c.jpeg.json","f690d9f88c5b7ec8a044254ff7949f6e.jpeg.json","f692676863cc9c9e570d6295e4af2399.png.json","f69f28589b5054d716184203b82fba8a.png.json","f69f3b6b7b6677a83f4baacf5ce7e143.jpeg.json","f6e5b7b6a9c9f9ccfeee2a074a0c1203.jpeg.json","f743c161a800793704d38252219216ce.jpeg.json","f792a9a7c74112b35edf7047f40f1790.jpeg.json","f7a32d5c00985cee6a728fd80f51955a.jpeg.json","f7c008134f887ca820e1c8d689038f2c.jpeg.json","f7d9c395b3900300c075fdf81f2312d1.jpeg.json","f7e8ea6ced92054dea82faebbe3fa243.jpeg.json","f80a0c5b5232ead4c9b22a01c17e2564.jpeg.json","f8326f743b6358281f3e0d6660e7f0ef.jpeg.json","f8f8c7d3d03bf90018b2666bf213077b.jpeg.json","f96bd76914f6f1ae7531f29c42a57eef.jpeg.json","fa26736dbe66222fdded414327ada3a3.jpeg.json","fa614d794e63113c22a47b5565423f9c.jpeg.json","fad737b5bd14428d8df30c7ac075b522.jpeg.json","fb17166c127c726b7a4d83f39860e307.jpeg.json","fb33c6ab0b31786fda22e2dfcca19533.jpeg.json","fb6e4592d5509d6a7e4d8b2809a883cd.jpeg.json","fb9423fb2527a12f984ed02a6b4fa0f0.jpeg.json","fbab704d149fca7dd1dd0d51ca8d48f2.jpeg.json","fbae3e2b20b46023ee996ca28ba376b5.jpeg.json","fbb044dbb5d3a328c9cd3f7bbd643561.jpeg.json","fbb1def90cf2d501e46d65e3a28c2c06.jpeg.json","fbbcddc555e1c1fcca7b2e427642fc43.png.json","fbd3b2c5b3c8ba8942a4e6c8bc32bf12.jpeg.json","fc0b12f2f06e054d3261db3932601265.jpeg.json","fc6bc2eb21708b3c8d645f709ef63927.jpeg.json","fc94c5366b7bdb40160f71f6d59ed727.jpeg.json","fd859c72c1a38cb382494882b1c502cb.jpeg.json","fda53bb004e548379f0296b9dc9769f1.jpeg.json","fe95e2167be353aac3ce9dae1c522784.jpeg.json","feb176b3f242f9f05b4fd137044be85a.jpeg.json","fecf9c5d753e1c29ce4dbad456290265.jpeg.json","ff548c8e48c6f3a687e6f406a7d0e47a.jpeg.json","ff71eb172eba1019e7cb579625089669.jpeg.json","ffb4e6b33755f68cc0ca9b93aa3af654.jpeg.json","ffe6f1ab63d429fa0e090ae78fb79301.jpeg.json","6ff8b403d437ec9210aed005f4b7d5de.jpeg.json","701a7573675975ffa17b5958d994da04.jpeg.json","70305901c9438c83241ded45a614a05a.jpeg.json","703dbe721760333b14f219b071af0258.jpeg.json","70644f2d900f3e3df8a094514378c8dc.jpeg.json","707c483fbfa6e10ea272052dbac91fbf.jpeg.json","70a226ff7b913d022f432a6262c432d5.jpeg.json","7103f3c81d7b9b5f47607f5517aa3210.jpeg.json","71040961f9178cb641ee2bda77daa1f1.jpeg.json","7153af94856dc09d65ca940b407ae437.jpeg.json","715f772a6a2306fd040598065dbe80e8.jpeg.json","71ec70fdebfe01fc82c055135e73e5aa.jpeg.json","71f38f7c6fd203dd193725a6575eb84f.jpeg.json","72058ea0b2f0ea32f6c1534dd0f4259a.jpeg.json","73203f4c115da50ef1bc411a2a043d22.jpeg.json","748311686961c891f4afbd8a373c02e6.jpeg.json","7531f8ba7c212c6ac13bdb1d6fc2a19f.jpeg.json","75349eb64fdd5d0fbf005df5c7594f07.jpeg.json","753681fbf3d3b2808b8c889f215feb5c.jpeg.json","755cb42ce7f1df46a1b893ea5d26cf26.jpeg.json","75f27276f628e427a2bf80aa7f9792e3.jpeg.json","75f2ace4925a2a749d6942885766f036.jpeg.json","76016b537ef69992905dad5c40289e21.jpeg.json","7624165ba78d723987fd7734ba6af201.png.json","76490fbac3548a57ce43fed12b29d5c6.jpeg.json","770ce887b2dc0c854d100ce8f11d30f7.png.json","77e3e59d693d7071abc68c4428802bbf.jpeg.json","7803ba050aa011496bb49f45ae395617.jpeg.json","787fd9830b9534d4603c704921312dc5.jpeg.json","78aa83fb0a9cd9148258af5201be89ec.jpeg.json","78b8ccc845900d7795b47d08618a0cd6.jpeg.json","78d5b85760cc495385c47f9ae486b37a.jpeg.json","78f3dad6dd4ace4ce16c1520c0c79a1a.jpeg.json","7941a02b69f84c5fe9faaa94510b2860.jpeg.json","798fd92d7074c29049fc772bdcb371ab.jpeg.json","79a1d77dd1380dbd15299ed72dfde5de.jpeg.json","79b0035db4d98d20736f354de4b7afc9.jpeg.json","79d6ffadb928a9aef20f1d3cd68810d6.jpeg.json","79f2078ec3fa4131062ec733c8d3ca10.jpeg.json","7a364f7e14fe9106787773b4bcb5cfc0.jpeg.json","7a6b1f70f5bd8e18f18c0c17a92848d3.jpeg.json","7b68ecb287ac3fdc8af59ed8d543848f.png.json","7b741584c0f339e91ecb7413c306d084.jpeg.json","7b7996d595b2df7ef2fd1a5502bb83cc.jpg.json","7b90dc94a75f26bd402735b26f45d017.jpeg.json","7bdc2b31b0632aff9922497b144972b3.png.json","7bdf2d802946995231b14274b2405ba7.jpeg.json","7c4c6e1c4cd562c7fc692fb3740dd1fb.jpg.json","7c5ce28f52c8befd9f86e6781a02b8c0.png.json","7cca72949282cc482fbb57b6230a92a7.png.json","7d3b0acd6672f0e36bd70153e53a07a8.jpeg.json","7d3d5a022ad59550f3ef259236e826bd.jpeg.json","7d514ff53e1150d4c1744aac4a2009ff.jpeg.json","7d52f80060a528bb7b2544c04baa65fb.jpeg.json","7d78e2881e3a2f454ce2cf822ca3e6af.jpeg.json","7e0ddfcc80d573770cd39f5853ec8daa.jpeg.json","7e27f9f4225299a55f786ab6d8eed544.jpeg.json","7e6296f2e01c6211eb7c5ba74ac3f4f7.jpeg.json","7e68577d4a092b66b009fe16f89b93ee.jpeg.json","7e9247a43ee9469014d72cfb7c6d23b9.jpeg.json","7eb014c79aa03c18f11a58ea58612c19.jpeg.json","7f9b22a160f257d7b4d3a80c09bb38ac.jpeg.json","806a1e4b1ed5a58f6b541fddad7f6a21.jpeg.json","8071807fba2c90fd85714cb40048c71f.jpeg.json","80787772d05ff9f8e58fc3e276df2c8f.jpeg.json","80a32fbe55da9c650fad4211badb09dc.jpeg.json","80ec0adfb26fc2d9e702382685560682.jpeg.json","8144caa0d00ba83634057866be21bb43.jpeg.json","818da784c33a36dade6ddbb4acb62a96.jpeg.json","81b0e7e46da0e150b295ee7a6d3e1d4f.jpeg.json","81e4ba2ee066ff09ce661cd1d9750ef2.jpeg.json","8202cf2736f818a34875d753d326b2e9.jpeg.json","820e93571774600596cbf7d152c12bb5.jpeg.json","828dbd12d2ed9d0ab59c39e0cf52405d.jpeg.json","82d4d7b29cdf9d02883002309f0a2b51.jpeg.json","831b1a75dbf1f605ac934cac47f2f6d8.jpeg.json","834bf7852cb1b14d1e1a3ba2a218bcda.jpeg.json","83631e520695495b92607c43168db581.jpeg.json","83c35f6fdf42ab14105836ede211a16c.jpeg.json","83e41860c05866c0e3918652417f8e45.jpeg.json","83fa5b54a47101cf0f63190eb24e1cd1.jpeg.json","841e1d2566656bded15d5e88c57c030b.jpeg.json","8441ce1f48fb9f42c1dc1965cc8c321f.JPEG.json","8475b7301e3b286871edd56452fd8e3f.jpeg.json","84914d767fffb2ba4fbeb2e04bf59d0b.jpeg.json","8494de96ec43cec4d3d2c33db672f403.jpeg.json","849ba4e71a31608b8dfa3407f50bf9a5.jpeg.json","854b6eb006aedb7c6ce72bdf5a078d51.jpeg.json","85838401549e4a1f22555627f44cc955.jpeg.json","859df8e5277a1c70815df5a5ca9ecffb.jpeg.json","85f59c64d02330962fbcfa26d7abfef7.jpeg.json","86040943da555695cf999b4ab9d03bb0.jpeg.json","866f7191af186f97a1b8c9623d73a82b.jpeg.json","86854f0064b17b918f90bc57a1980437.jpeg.json","87b431be307a704329f32ee4a63ff10d.jpeg.json","87ca351b10e672fbb1b3dde77e5525e4.jpeg.json","87df1901f592db9f071adb357b3c2d42.jpeg.json","884326038ea0d8ab9a9aa7847a87b73b.jpeg.json","886ad98d2d123c4192efdc0a00df4635.jpeg.json","888a1119743c85ff9120870f1c10fa53.jpeg.json","88badd516a416a5dd1528759d81dfa2f.jpeg.json","892e21ec3aa8c1ed1765a431b2b68482.jpeg.json","8976d9fc2c5b2ababda1406d043bdb74.jpeg.json","8a4e32e64a7dd08e29804623d4b96547.jpeg.json","8a59d79dae544c9046fd12f476c97165.jpeg.json","8b4bd50543e08290446d32a2a12f6ffe.jpeg.json","8b56fb59e1e1fd22a2df2013dd438c35.jpeg.json","8ba2afe0a859ca8156b8310badca4c9f.jpeg.json","8bb8ab40226c6368f143bf29c1bbd5a6.jpeg.json","8bdd8a6919661ea6cc729a31ab0955d5.png.json","8c3b0c73ce253840febcd51f3b8e8ce6.jpeg.json","8c5f38f5ab076909e00b447248d00687.jpeg.json","8c9496dc6cef4a0983fb3973b55b02a3.jpeg.json","8cb75fbf0624fd5cd086a49b1be8b2bf.jpeg.json","8cbe52a94b0701078098008e8f95c6ed.jpeg.json","8ce2b5a1afefdcbdd3462813a55090df.jpeg.json","8d0a85407a1eb4ed4162030d482275f9.jpeg.json","8da7129adbc4cb11a0387d3542fcf615.jpeg.json","8db1f21462137313d55f1a162ae1dd76.jpeg.json","8ddba860f36f70d82496d0f9d0b88bf1.jpeg.json","8df2cacbafde9eb289e229ea80f156f9.jpeg.json","8e2a21e0bf7fcb73adebe3dc760aac3b.jpeg.json","8ec4f340ff26d4d3685d55a9e51664a2.jpeg.json","8ee0a6ee1d87e42c377ff2b96c282f02.jpeg.json","8ee56a12c1f8929e0af9ed466e798381.jpeg.json","8f06cd4ff26dec9e56ccd07220446075.jpeg.json","8f0e47302b522ed20a7520c5b4a78f7c.jpeg.json","8f618dfa969081acd3012b296e2201bb.jpeg.json","8f9b1e321b35dba0d6942180979cb66d.jpg.json","8fb501f2238f9eb0d2d1026758cd15fe.jpeg.json","901c453cd81a877ace73eb0ad746a2ed.jpeg.json","9032a049b9745b29b5435ed253a49ff5.jpeg.json","90a2d791c096a54f78f0b8a04bbe8da2.jpeg.json","90a7881307bd13eb4bc9be47aeb20f28.jpeg.json","90c4545574be9de286262b7eff20adac.jpeg.json","911f4132ffbd482d9eb7f85f66b2ecee.jpeg.json","912f85ed5abd31fb7adb24c2bd8859d6.jpeg.json","91b1be0ae205888ac0b77ea6320e2ca1.jpeg.json","91d9f5ab78276a70f0972e3d8a15f745.jpeg.json","91db4d6a662b759a72d48707ba43e34b.jpeg.json","91ea2f5359cc19280d96c101bfe22759.jpeg.json","921a572517b61ec2d8096256b22fbeb2.jpeg.json","9236e75942f85a5a014e93aac73dfeff.jpeg.json","927810f080a7df039fb429fc1034c146.jpeg.json","92d8769ab1ac7919e46086e267349ece.jpeg.json","93f5fa9beb051efc61b931463b3362a6.jpeg.json","941afb668b7b5432667fc0cccfbfb641.png.json","941c9453c25b97cf56b81daf0fa3e2f1.jpeg.json","943cb181858d995be5225d1419e50c5c.jpeg.json","9441715282a08469d681a97ccd4b233f.png.json","947ddefa87430eaf0c40729d0fc7410f.jpeg.json","95355a69b7b39c60259719705cea9900.jpeg.json","9558b20aaeb1b1c833e20ccdf413fbf7.jpeg.json","957cad737d5e939581f3e8aa027233e1.jpeg.json","95c563f18662d1987996cdb2ead4624d.jpeg.json","95f5e3159aaebf5831ef0409cf274075.jpeg.json","9639be2d1eff02ec7162a823e669c44f.png.json","963aef3fff65546cb9803cfaf889ca34.jpeg.json","96440c01c37a3a594dd42ef21431f3a9.jpeg.json","96450269ec8e68b0108e8b5b4d145c55.jpeg.json","965c8a1ffbe3fc9951ee1ef5e34c6d37.jpeg.json","96ab72a688f57d0de88c71ef0bb6cfe7.jpeg.json","97738f79dad5424d6e1a440a3a461532.jpeg.json","97b8f38e9cf3adbf6657b25ce1a7335c.jpeg.json","9813582798fab064746a1b94a0f38f24.jpeg.json","98ccbfbe45bc1bcfbc3f40ec39a5f41f.jpeg.json","98f688756e56414609e0e27d39701c56.jpeg.json","99366e7ee9249b7dff69856eea8bd9a1.jpeg.json","993834e21792dfc43bfa9dd6bac01013.jpeg.json","997c7bd5d32ec45ea86a13b11444a59e.jpeg.json","99e22017232b6de400542df046a784ce.jpeg.json","99efc5d7227e94e8fa5db14657440d93.jpeg.json","9a41ce7c515fca5f0e3d852553bf7401.jpeg.json","9a586f9a33601864bd887e6a3fc15306.jpeg.json","9a9adfe19d6c60cf14406ee7cc4ba8b9.jpeg.json","9af6d1a0b444caef74043671fecc5a2d.jpeg.json","9b01e5e1794f2049e77f62be7d61fefb.jpeg.json","9b09ec48138fcc10bb6546e561550205.jpeg.json","9b71abcaaf16fb9544330871c1997c61.jpeg.json","9b72b7e9c7a2eba1225a2b94bf6c075a.jpeg.json","9b74584ba01059866029f10e04a28d86.jpeg.json","9b9f0306056eba83ee63fe13309d6416.jpeg.json","9bc01accf2bd998927e35eac49fbe701.jpeg.json","9bc603c6f4135cd85aadfb0ce280cf2e.jpeg.json","9c11c8da895be5e51f6f376bd565a19f.jpeg.json","9c1fa32f357d296e353e8eaa1c0de5d5.jpeg.json","9c417c45c7c2f840657b66c5c6445e1f.jpeg.json","9cfe50b051a74f0d34d47b2c2fc289ff.jpeg.json","9d833f506e39a9b256b7895ca4f4a12e.jpeg.json","9e6788e49029872d6af65628e47dfb42.jpeg.json","9e753701f51917034425ec42089227d1.jpeg.json","9e94808c0f99648e03ca95de2db13aaf.jpeg.json","9ed3aeff287805318f9158e973a84d80.jpeg.json","9eec8bf302a0b554a97ec051326e207d.jpeg.json","9f0354346e5f10f3a9cfa52298f988af.jpeg.json","9f470ac5164bde63ef3be45e9710e55f.jpeg.json","9f814896cba97063fe41f14519f3e3a9.jpeg.json","9fa8ba16bdb2e9866dc88edb3d6d4dbe.jpeg.json","9fb879fc041e93f09f4d4220f54ac496.jpeg.json","9fc6ace3668c28d8843ce3faeb967a81.jpeg.json","a11192978ca202719cd557e43ed49985.jpeg.json","a2985d563eeb38e0cd9853d9874d2186.jpeg.json","a2a3ec05647e4708b65b988b72831a99.jpeg.json","a2c345b97444160ca424f69ab2967424.jpeg.json","a2e6f5eb92c42c6c081375193e4016c0.jpeg.json","a3078d34dedd5f24001b5e0c629e53d1.jpeg.json","a36349468a0f3303e247c3bf1a3624f9.jpeg.json","a36cc42841175f14579ff80d9c21c7f8.jpeg.json","a375ca2f02bce027332edc0b59412819.jpeg.json","a43687a677c502eda3825a8e22504865.jpeg.json","a4a36933aa80f535d0d26beef6d534df.jpeg.json","a4fdc6f8ff15a377a7bbef763ce056b4.jpeg.json","a50e6b1510a72ee28a09f8f89201a515.jpeg.json","a51cf5fbcebabdb1b9b108349ef4c6e5.png.json","a56531bd2a7405fa41ec2c89a7740772.jpeg.json","a56e1286a9516e4c2a1ef113ab6b7f9c.jpeg.json","a6284ea25b31c6d9744e30e430ba22bb.jpeg.json","a64bbd93b3c31558166c74267c4ac73e.jpeg.json","a701b6c44062f5392e9375b3a70211a1.jpeg.json","a7582f2944bf8b2dbffb613d5e81911e.png.json","a768268f66f7ef309cb5adce6932e916.jpeg.json","a84850991d22f6dbcdac714225d2ed5d.jpeg.json","a9168a323a9861c5d44a27ca12cdcba3.jpeg.json","a99c7bbde59c831006c02ba047b3895b.png.json","a9a3c66d69397aabc475dea5b3d46fd7.jpeg.json","a9c70118fa87e20076152f8df3303e0b.jpeg.json","a9d60d9edf124b2a6e0371a9d4b1105a.jpeg.json","aa21c7f303b2b5e161883704901ca315.jpeg.json","aa248053b9f210e59cebe829e3c9b7d5.jpeg.json","aad7d6917fb9297e3647719963062f23.jpeg.json","ab10e5cb60fe670021c78c78adbab5df.jpeg.json","ab613048422816fdf3c15608065a29a1.jpeg.json","aba73d12fbe9f8e1b5477a197e07dfb3.jpeg.json","abdd45bcbb22860d36069b0019247198.jpeg.json","ac1743e4e6991795d887cdf34e7f68da.jpeg.json","ac294fcb10136cc069ca76fea18b1012.jpeg.json","acd9f38ee4702a6539408c69f593e091.png.json","acee560d56cf1e1c3be2fe5cbce85311.jpeg.json","acf27a1a77b1c64826d23fdd8b314de4.jpeg.json","ad8afd075c3ba5a3ffccb866223f5edb.jpeg.json","adcfa03f6ca674ae5c597dc120e76a89.jpeg.json","ae35ae8e2be13b5b45595c274b1cf7eb.jpeg.json","ae452077f9971d9921a7afb3922048cd.jpeg.json","ae6df2a0f14877bc5461abc3340e50c8.jpeg.json","ae6fc6f56c8727760e79d8c2b68036f2.jpeg.json","ae6fe194927e756df7fe631522f1c610.jpeg.json","aef1eeb4fd79ccf182859a3246c1d8d6.jpeg.json","af201f933bac670b2dc2350e86cee5c1.jpeg.json","af3f7fde9d22f417cdc0df4753e1cb50.jpeg.json","af7a777c97af1ad15984c4a5301ebf33.jpeg.json","afe3e53779e69ca42aab7f1791a5da39.jpeg.json","afe5047af1f74252f2c9c60992b91056.jpeg.json","b009a59ff072a43ea399a6e5ec517abf.jpeg.json","b099b4fb9dbe29bfa7c7b6be1940b614.jpeg.json","b0d92bfe4af109347d4463c46239d39e.jpeg.json","b1507b84f010a6a9422dfc64fb543487.jpeg.json","b16ed70eed5fb43186894fe692e7d475.jpeg.json","b29f34f5ae0a0809fcfa977b5ad9da87.jpeg.json","b2b7b90e86c5825806925d455e5bb216.jpeg.json","b34a507a4e4cfb5ebeb22839e3720d9f.jpeg.json","b369335407124257410680bc75a29766.jpeg.json","b3732c175eb8b6f9cd1ae9045166e16f.jpeg.json","b3e5dfed31c4e7bc8effbacf61ff82c8.jpeg.json","b40755769b8c56c01dd68b0dea7845cc.jpeg.json","b42335a9da636e4ad3cc00d3cc4ba05d.jpeg.json","b426df0cc4acc34ec5af159e8d42efd0.jpeg.json","b43941538c97ed7b14e4b5247f791e41.jpeg.json","b449be070cb8539436be650255bda41f.jpeg.json","b4a8395a54e094ee1250860e34c2ec07.jpeg.json","b685e3383780ffeacb8e6297cf0ae4b3.jpeg.json","b685e5f2be5e93d34e8041c66c1f1e03.jpeg.json","b6d011b0bb1a62abaab18b468449f784.jpeg.json","b6d427bdcb4ccd66b43f2bcfeea2192b.jpeg.json","b6ee014f5d623c6cca88732ac6d41c7c.jpeg.json","b71b3c9a493a1f63adb497cbf6cbfb2a.jpeg.json","b7355aa0b18badfff95d5fd5af6b8268.jpeg.json","b7616310f7f218dd57906e6e0cfcafa5.jpeg.json","b77a8535c138319408a0c757fffb1d79.jpeg.json","b8688ea021519777f61f124a419fdb29.jpeg.json","b86b50c9f0ce0b64431c4081a2f573dd.jpeg.json","b8ae48811862229e89f88794a2f02ddc.jpeg.json","b8cf36cfc7f9ae5a48714e475b25f164.jpeg.json","b945150c2ec585f3d7bd0a9102f8540d.jpeg.json","b95079e842b08ff260630efe2fe7e141.jpeg.json","b99f523d94bb7df35b7d028fdf94e988.jpeg.json","b9ccf86f5471af4d6428f873d8d958a8.jpeg.json","b9dc2bea6f9241ebeb5fc7e7ff8c04c6.png.json","ba1071f184a84439df9f6114b6e70472.jpeg.json","bae4b514eed98090d7e9aa81e0b03de2.jpeg.json","bb41fb0ae0d6f9a036a56def9924f172.jpeg.json","bb8648702ffaba9723ce164f24ca16d2.jpeg.json","bb8e78ad2c05150f55ce85704c4023e5.jpeg.json","bb8ec58ee700b811a205f577450160f1.jpeg.json","bbbef3b7c7943694c0969e11c35179ad.jpeg.json","bc390abe92c933d67427607a90c5d8e0.jpeg.json","bc3c5a9d4543ed7f5795582a632b20ce.jpeg.json","bc437be6d3ee896e081a752a7f7dd9f1.jpeg.json","bc6a12c985b8ee2d74d4c9bfb9918791.jpeg.json","bcee2f70904deff9fbfa47186c382dab.jpeg.json","bd78ced66d4e9939f7c4dca321c5cc64.jpeg.json","bd88325d1e6c52f87981f7b88d9f4efe.jpeg.json","bd92ceabd72484a556792c0efefc9071.jpeg.json","bde5e4571fee9681d6e7f2e089b5e4c5.jpeg.json","be26e439c5348cbcf8802fb66ea10ade.jpeg.json","be5c23545eb93575e7060f01f7ddcf88.jpeg.json","bf98d5ffa7eece6b3e72683e29c184f7.jpeg.json","bff24593f4c61fd4cfe25a63f019822c.jpeg.json","bffbb92b56845b871a727abb6e0cc9b7.jpeg.json","c03a255bb791b438134bcc211bcfc229.jpeg.json","c064de930c0c79937e528b0142a30da6.JPEG.json","c097e31ab213847094af00c163c27223.jpeg.json","c0bc62c518d17b377e60b8c6c046b862.jpeg.json","c1505ff920b8712d5d221e9ffe70709b.jpeg.json","c17e47a687117911eaaab9eca760c4eb.jpeg.json","c17fab603ec0c74155b1636c4418696d.jpeg.json","c1b15b94b3dfb7c964cc0027c8258ee7.jpeg.json","c1eb726692d48e56505f69d1d5e8f93e.jpeg.json","c238aea63cbddfbd121c944efc62af6b.jpeg.json","c25288d3b882fc081156387819203347.jpeg.json","c2e97bd83465c1013fb449aae9c139ba.jpeg.json","c311ca4d933f354d9a1b7e1a0ce5a158.jpeg.json","c37fcd546f12a1027b3aaa98ca49b567.jpeg.json","c4709b17ce7ae094ef8d148567aafc12.jpeg.json","c4dd7333fd768946314126dbaa209044.jpeg.json","c5428858d364875f454de16ea20209ef.jpeg.json","c54ae5db16c59a7689c3181b794776c5.jpeg.json","c57c739aa8fdffc25b3df14cb5e8cdf2.jpeg.json","c5c9c9f3078b29003c3eed19a9043419.jpeg.json","c6aca7bf440fd6f83df844419739ff34.jpeg.json","c6bf4ff2724a16eba0c61405e8739b03.jpeg.json","c6f987545f6ac44a7fae4b0d90c1cc22.jpeg.json","c763aefb9189781ee2fb49768f06e2e8.jpeg.json","c7988e0dbf6876b324824d8d58103fba.jpeg.json","c7a5775839e1c03994860dc36bac13f4.jpeg.json","c7ff714283d234f96601bcfd20c3ec03.jpeg.json","c839ef18856b89f21e79652fc322b936.jpeg.json","c862eaddd35a6123ce4eb3af530525b3.jpeg.json","c8702d40118aeae6ae328fe2b90937e0.jpeg.json","c8cc592e53418976b7d906ea53a20bcd.jpeg.json","c92dc36627b6ef845c0430b314baa1b5.jpeg.json","c9452a4f57b646744407a8d6f01e5e59.jpeg.json","c94ff10637406f0aa3c904f9913a7fbd.jpeg.json","c957f1c5f4873e6c3654e0e859ca2887.jpeg.json","c9af4e35587383ea25a9bebb1fc7b90e.jpeg.json","c9d7d4a01962409f8660829649a266aa.jpeg.json","c9fc3d14fe098f8a60e83ecf05072f8b.jpeg.json","cae0f91f1011966b709c653763e689cd.jpeg.json","cb579a43c90f5ac0c067523dc3a4d439.jpeg.json","cbaa4ba0460b5b82aafeed1065407d55.jpeg.json","cbbdd74c5e5ae3ee1ca65d78e1aae440.jpeg.json","cc312bc2f7d212f7ac16d680a08d132f.jpeg.json","cc3f325ba4ac7e580a93a7a21df57ddd.png.json","ccdc1da1cca2d2037f4fef94fc043853.jpeg.json","ccdc67e21001c5e546dcac8449598539.jpeg.json","cce9be1ce40f81e773a4e881ffc86470.jpeg.json","ccf563a0e70c62e3208854ef7c5b827f.jpeg.json","cd2d43a111837b154397252ae669f3fb.jpeg.json","ce382ad55ec548e681cf4a04a59a5580.jpeg.json","ce3863ca96e401fbba5adf438d98756f.jpeg.json","ce5b568520be24a630d7d6a7f9e4bf80.jpeg.json","ce65a1854e71ea6c4bb7416088713ebd.png.json","ce7318afda1c720cfa8808d4a4b512f0.jpeg.json","ce8110205ea28a8e18aaea50ac566fc7.jpeg.json","ce8519acff3b6c7877ffd964384ba7bc.jpeg.json","d03f62e47e2cc64f827d8dc3c9a56882.jpeg.json","d08e0703342cadb0cb4aa2251afd601b.jpeg.json","d0b07855a47678b6a8d7ce4552771c7c.jpeg.json","d0c5622bc33161aa3721cd7ded296c9d.jpeg.json","d0f3c17be64e044984b83b9e4aee4e3e.jpeg.json","d114f36c25a348f912acc7aed1f05d23.jpeg.json","d133c00b5989c9b453dc30b78cd9552d.jpeg.json","d148850ed7307574db17f9d965a2bb90.jpeg.json","d16a63cf6ba1a70d910a4c6689952a7e.jpeg.json","d18f1718fdc29baf281eac12b3a18b3e.jpeg.json","d1b4fdd0fe09e829944f8dad6efb0f4c.png.json","d1b72e014226de6f9b9f344e02860830.jpeg.json","d2130727129386cab1ed53a87a56e239.jpeg.json","d2e4e8808a62c404f38498faf3b86abb.jpeg.json","d310d9ddb154fbeeece93e5e85940bbd.jpeg.json","d334da1da9ff6e2975664b8005e265db.jpeg.json","d3ad82ac8936c1a68a3da54f7bd4217b.jpeg.json","d3c482aa1c424068f5e4244022b0ee49.jpeg.json","d3edfe6bab836691ec79f363a5c796f3.jpeg.json","d3f7d6da6e9ef8fd6dbbeff7b05afeaa.jpeg.json","d401f57bc3aff32e363b3168dfa86e62.jpeg.json","d44d401b12427837b14c734be51e0aee.jpeg.json","d467ab7b3d8beff8d51024d3f6ccf6c3.jpeg.json","d49ec827d0b340989f64be6fc6f996f4.jpeg.json","d4bc6f134ef7d2e087be9cf908c2a2e6.jpeg.json","d576adbe497d0859636e9362565cccb0.jpeg.json","d585bcb7110dbbd3d2c448fbd56e1de4.jpeg.json","d5af05d60ec4ecc0bfb110f58b175ab0.jpeg.json","d6330df9b0f62cc9a66b5bfc8bbc9949.jpeg.json","d657eb1e8aadf3c89cba389a6c118ab3.jpeg.json","d6913dd3374d981053042e99699aecde.jpeg.json","d69600cc178936838fb7fdb1118e7537.jpeg.json","d6a7b438e7dae8b357012f006a5853b9.jpeg.json","d78df2ab61897b99f3a08831ec348e45.jpeg.json","d7aa773a4dffafe161c0e30de3379324.jpeg.json","d802cfd3bbbc3276258ba689c6999a32.jpeg.json","d81e16b7e522222cd27a51b483c11800.jpeg.json","d87a53001c3cc399b76a638f731d2cc7.jpeg.json","d8ed135ab9d50e136394a2256a6c9d46.jpeg.json","d8f8f0507a48b76ef59060bfa9b2cf47.jpeg.json","d915c4b42c400f9bbb7ed5a713375e86.jpeg.json","d931f074be18d83bc76fad43dc165081.jpeg.json","d9573cbf00349536974d9ded8223e4c9.jpeg.json","d9c2b4df4f06e4715feff4f50a073a03.jpeg.json","d9e725406bee40d5acd6c8042031855e.jpeg.json","da14aaf7d6c01bc230a5d2c82d1c9500.jpeg.json","da233e2bad57b401c494a4f62883e34a.jpeg.json","da896edbf57d593dd39247b5a5965ca5.jpeg.json","daa446ba29c460532f352814f4026685.jpeg.json","db200ab7ac7cf86e7cdf86da6347edc5.jpeg.json","db3266e571b66e3ce100c0033710cf1e.jpeg.json","db61ad9e74d0f52204608b894dccd498.jpeg.json","db656485be6ccaa7163c9561ce8d4db0.png.json","dbbee0ed60f3d78f7f3d40c4e9f04b05.jpeg.json","dbcce7ec562b55974af45addcfcdb291.jpeg.json","dc1e1228d473d40b6e41712a5db17366.jpeg.json","dc2c97305ce4686d135c86fa48c204f0.jpeg.json","dc6f2a9dd6fc900c8b1e17c33eb9e2b1.jpeg.json","dd2a8c3d9757912e71a7b89466b0dbf9.jpeg.json","dd3c8b67685c6a59b876a3e01d393d13.jpeg.json","dd562c58de783991b44578cfcee09ee1.jpeg.json","dd5658d1decea90a9af7a37d1031f8c6.jpeg.json","ddc0ed9c03c92a1d89064132f8be0403.jpeg.json","ddef41bc01129efd1f84bb3555bca104.jpeg.json","dec14706cfee632134eb72fd3821db6b.jpeg.json","df0c46342a07183d8bc8c73fc7a8807c.jpeg.json","df2264b8511a4ceff8b708982e4cf4c7.jpeg.json","df529f96a1a27b98a412aa6507c967fa.jpeg.json","dfb5016b199c5c7a68af6fd54dff8a5a.jpeg.json","e049759552aadbb681faeea21a386ddc.jpeg.json","e065afa7e8354ff5d2fa6b8fc2e7133e.jpeg.json","e08cbdd522c38bc9622d4ac7055f2f13.jpeg.json","e09f9a594a5f082e1b5320dcc200f73a.jpeg.json","e0b13dae2ee6176ea0997a69c73edb6b.jpeg.json","e0edee1feb02e46a737d785731660606.jpeg.json","e1627a23c077a873c8b3feaa0865abe4.png.json","e1819aae68143e22365034525c067dc8.jpeg.json","e225981599523f56c082cef434378250.jpeg.json","e26b4a319348af33c3eccaae584fb273.jpeg.json","e2cbf82cd42642f86a79ad520cc1ff4c.jpeg.json","e33ebc36feb28144cc11e13a3d554625.jpeg.json","e352f13f8d13bb47833807e01f1ed4ed.jpeg.json","e3961ea3be6f5072723fa8a9b481ba35.jpeg.json","e39ec5e9d0761cb2f42b6fd951a8cdb1.jpeg.json","e3a62394af2093967325d9c48fe96e97.jpeg.json","e3e614d08c544a28ecb766e5dd2b0285.jpeg.json","e43ab595eb777ac412a9234a6fb289cb.jpeg.json","e45c0e2626ef36f047157c73dd8e043c.jpeg.json","e45cebee744e770e06250863575a61be.jpeg.json","e4fbf2a865ef90de7ba12b8a20b16ed3.jpeg.json","e506d47988a469e048f0d30c87a77269.jpeg.json","e51fef4ae33be241c98139b9e5473b0d.jpeg.json","e5578d64e93286b225f8889e57243537.jpeg.json","e5bedd47dcfbe72338ed8887326dfba0.jpeg.json","e5e536c80dd4645521d4b116a323d23b.jpeg.json","e6197415a8e971b2929c42c77eb935f0.jpeg.json","e64a0a52ca7510f358a66c826dd87703.jpeg.json","e66835e2dc2976f5f89c7b431a25b18a.jpg.json","e674947be8a06387107332df62f0f9bd.jpeg.json","e675b5f02a6ef6f054d915aa40b0376f.jpeg.json","e6a2e3f440439eea3d71efae67282fcc.jpeg.json","e6bcf7cf0fa9159a0cda27a29142de91.jpeg.json","e726e0061a9c28d232ad4a82075f9163.jpeg.json","e76165960c6acd2e981542adb051cb88.jpeg.json","e77a77e463d77b7fdd5493ba34e07ae1.jpeg.json","e786698b884afd2b4d2083f54752d13f.png.json","e78dba436cc506b491635587931efa45.jpeg.json","e7a000d9dbc511424ff1f4dba6c0d66e.jpeg.json","e7bdb14f10d4d024113702a7ca8661ee.jpeg.json","e81b86a127345c34837a0fd6faad4bb3.jpeg.json","e82a9debbbd86a16d08c7bb89d1868a7.jpeg.json","e89220fdf4a509c52ae62ec57ebeb8d4.jpeg.json","e9181e87b3fedf0371ab1d22f9e98cd7.jpeg.json","e9266d87658a25bcd0ee1a6000bf7c3c.jpeg.json","e94e1762fbdfee04b1b882cda844175e.jpeg.json","e95cc2a5919596c9411f30a5aa0533f2.jpeg.json","e970e9e2ad1ecffb03ea07bb847d244e.jpeg.json","e9b245358b4b5e1cd1dede8623db322f.jpeg.json","e9cc8329e12b6150a29bb24f7ff7baa7.jpeg.json","e9d8805725a216a4e90bc68abffee7f7.jpg.json","ea1c9a1224b1ebf21bfbcb183041967e.jpeg.json","ea5164f1625fac7fc25785146105026e.jpeg.json","ea5255de62e00538f7a5552d95ffacd2.jpeg.json","ea848887a2d74fe043f81f56b88a2d12.jpeg.json","eaad3399669a44432d88bcefb80dddea.jpeg.json","eab3e34e6ca878e5b16dac1d1e1a6dfc.jpeg.json","eb8c79d310d7054d37890d2290b85bf0.jpeg.json","ebc8c16eb8ebefa11516f6d9b2990edd.jpeg.json","ec23245f92b5a1bff7db98bac02a6cb4.jpeg.json","ece1ca7a319c25f3cfeec23fa97d2fc3.jpeg.json","ed243bb698dc4b19ce281a770034f9a5.jpeg.json","ed4f181c6c8863f2ded095700bc2c7c3.jpeg.json","ed6409e995cff02a1f5a30a024326aef.jpeg.json","edb4ec033232cc5560d21d1a91028ac6.jpeg.json","edb5618bb7c3ce70faa9cea170958efc.jpeg.json","ede4e9ca14dbb92cdaf95576103ac6d1.jpeg.json","ee08415d8550a2d87364de650a07ecd9.jpeg.json","ee6399df5a80d9cc28c7b13cec0abd1f.jpeg.json","ee819fb1e536987c8c566f62791ef8d9.jpeg.json","eea87fd1a1ddc013631567ad9bd670f8.jpeg.json","eecb407083fd3cf5c704f6b0d86ca8c8.png.json","ef3737f8de72aeb3e967b8344be0f5f0.jpeg.json","f0223dc36cfb60e12b176050796453ba.jpeg.json","f0c7529b755b21c1c7f775413203e49a.jpeg.json","f0e8b78f1f18daf76dd4fc9d04211720.jpeg.json","f11090e810ba8d57814a572d78d40b56.jpeg.json","f17b12b2416962a2d668c8899ff2f0b7.jpeg.json","f18e6c5d2c65873cf38884f41489bc0e.jpeg.json","f18f8f3e6d3d18bc98b0818ee43b32d8.jpeg.json","f1d0d1ebe22add8e060e73cf30048f54.jpeg.json","f1f48ffb9b03fdfadb469dd198aba7ff.jpeg.json","f256210bf91062ad03c442bd67e8ae5c.jpeg.json","f264ae3144218ac7e01b5c9f8c8ead0d.jpeg.json","f2cc05d2d2d65f4fd123fc30197d94d3.jpeg.json","f2db964b5fc68cef96db0381a3b3ba77.jpeg.json","f2edefb4489da7af9e3db508a4cdad48.jpeg.json","f338b2e04ed7dd590a786fa7a0466b3c.jpeg.json","f344d8a79230d6eaa5a79d6fe937807e.jpeg.json","f3460f571c21ba0024630a81f7bfd646.jpeg.json","f3c689b454a7ec1c0d4a656e5c73627f.jpeg.json","f4137bf001ab4487db97aa7f8c02f02c.jpeg.json","f485d8be1fa45e835d1180975a32cf0f.jpeg.json","f4d42e49993c2ea60788c5080b8bfd60.jpeg.json","f4f73b8baef2980de373858ae764234b.jpeg.json","f550e853f302025076d54ebd00ce5be9.jpg.json","f5604bfa425414845c84cbdc5f374d01.jpeg.json","f5abe1eac1b1a79fa194e6a6afec45fc.jpeg.json","f5d8576e4c189d11d057a6412dc81171.jpeg.json","f652d7303b131cd7531f1b13e8aca173.jpeg.json","f68d939fda43324776d14209edcee17c.jpeg.json","f690d9f88c5b7ec8a044254ff7949f6e.jpeg.json","f692676863cc9c9e570d6295e4af2399.png.json","f69f28589b5054d716184203b82fba8a.png.json","f69f3b6b7b6677a83f4baacf5ce7e143.jpeg.json","f6e5b7b6a9c9f9ccfeee2a074a0c1203.jpeg.json","f743c161a800793704d38252219216ce.jpeg.json","f792a9a7c74112b35edf7047f40f1790.jpeg.json","f7a32d5c00985cee6a728fd80f51955a.jpeg.json","f7c008134f887ca820e1c8d689038f2c.jpeg.json","f7d9c395b3900300c075fdf81f2312d1.jpeg.json","f7e8ea6ced92054dea82faebbe3fa243.jpeg.json","f80a0c5b5232ead4c9b22a01c17e2564.jpeg.json","f8326f743b6358281f3e0d6660e7f0ef.jpeg.json","f8f8c7d3d03bf90018b2666bf213077b.jpeg.json","f96bd76914f6f1ae7531f29c42a57eef.jpeg.json","fa26736dbe66222fdded414327ada3a3.jpeg.json","fa614d794e63113c22a47b5565423f9c.jpeg.json","fad737b5bd14428d8df30c7ac075b522.jpeg.json","fb17166c127c726b7a4d83f39860e307.jpeg.json","fb33c6ab0b31786fda22e2dfcca19533.jpeg.json","fb6e4592d5509d6a7e4d8b2809a883cd.jpeg.json","fb9423fb2527a12f984ed02a6b4fa0f0.jpeg.json","fbab704d149fca7dd1dd0d51ca8d48f2.jpeg.json","fbae3e2b20b46023ee996ca28ba376b5.jpeg.json","fbb044dbb5d3a328c9cd3f7bbd643561.jpeg.json","fbb1def90cf2d501e46d65e3a28c2c06.jpeg.json","fbbcddc555e1c1fcca7b2e427642fc43.png.json","fbd3b2c5b3c8ba8942a4e6c8bc32bf12.jpeg.json","fc0b12f2f06e054d3261db3932601265.jpeg.json","fc6bc2eb21708b3c8d645f709ef63927.jpeg.json","fc94c5366b7bdb40160f71f6d59ed727.jpeg.json","fd859c72c1a38cb382494882b1c502cb.jpeg.json","fda53bb004e548379f0296b9dc9769f1.jpeg.json","fe95e2167be353aac3ce9dae1c522784.jpeg.json","feb176b3f242f9f05b4fd137044be85a.jpeg.json","fecf9c5d753e1c29ce4dbad456290265.jpeg.json","ff548c8e48c6f3a687e6f406a7d0e47a.jpeg.json","ff71eb172eba1019e7cb579625089669.jpeg.json","ffb4e6b33755f68cc0ca9b93aa3af654.jpeg.json","ffe6f1ab63d429fa0e090ae78fb79301.jpeg.json"]
\ No newline at end of file

tools/train/tinyYolov2/test.html

-<!DOCTYPE html>
-<html>
-<head>
-  <script src="face-api.js"></script>
-  <script src="commons.js"></script>
-  <script src="trainUtils.js"></script>
-  <script src="loss.js"></script>
-  <script src="FileSaver.js"></script>
-  <link rel="stylesheet" href="styles.css">
-  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
-  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
-  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
-</head>
-<body>
-  <script>
-    window.tf = faceapi.tf
-
-    // hyper parameters
-    window.objectScale = 5
-    window.noObjectScale = 1
-    window.coordScale = 1
-
-    const weightsUrl = `/tmp/tmp__224_35060__320_41188__416_31050__608_16520.weights`
-    //const inputSizes = [160, 224, 320, 416]
-    const inputSizes = [512, 608]
-
-    async function loadNetWeights(uri) {
-      return new Float32Array(await (await fetch(uri)).arrayBuffer())
-    }
-
-    async function fetchFddbJson() {
-      return fetch('/fddb-detections.json').then(res => res.json())
-    }
-
-    async function run() {
-      window.fddbJson = await fetchFddbJson()
-
-      const weights = await loadNetWeights(weightsUrl)
-      window.net = new faceapi.TinyYolov2(true)
-      await window.net.load(weights)
-
-      await promiseSequential(inputSizes.map(inputSize => async () => {
-        await promiseSequential(window.fddbJson.map(({ filePath, rects }) => () => {
-          return test(filePath, rects, inputSize)
-        }))
-
-        const losses = Object.keys(window.lossMap[inputSize]).map(k => window.lossMap[inputSize][k])
-
-        const totalLoss = losses
-          .map(l => l.totalLoss)
-          .reduce((sum, l) => sum + l)
-        const avgLoss = totalLoss / losses.length
-
-        log(`totalLoss (${inputSize}): ${totalLoss}`)
-        log(`avgLoss (${inputSize}): ${avgLoss}`)
-
-        window.losses = window.losses || {}
-        window.losses[inputSize] = { totalLoss, avgLoss }
-      }))
-
-      console.log(window.losses)
-    }
-
-    async function test(fileUri, rects, inputSize) {
-      const img = await faceapi.bufferToImage(await fetchImage(fileUri))
-
-      const groundTruthBoxes = rects
-        .map(({ x, y, width, height }) => new faceapi.Rect(x, y, width, height))
-        .map(rect => rect.clipAtImageBorders(img.width, img.height))
-        .map(({ x, y, width, height }) => ({
-          x: x / img.width,
-          y: y / img.height,
-          width: width / img.width,
-          height: height / img.height,
-        }))
-
-      const { reshapedImgDims, paddings } = getPaddingsAndReshapedSize(img, inputSize)
-      const squareImg = imageToSquare(img)
-
-      const netInput = (await faceapi.toNetInput(squareImg)).managed()
-
-      const losses = tf.tidy(() => {
-        const outTensor = window.net.forwardInput(netInput, inputSize)
-
-        const {
-          noObjectLoss,
-          objectLoss,
-          coordLoss,
-          totalLoss
-        } = computeLoss(
-          outTensor,
-          groundTruthBoxes,
-          reshapedImgDims,
-          paddings
-        )
-
-        const losses = {
-          totalLoss: totalLoss.dataSync()[0],
-          noObjectLoss: noObjectLoss.dataSync()[0],
-          objectLoss: objectLoss.dataSync()[0],
-          coordLoss: coordLoss.dataSync()[0]
-        }
-
-        return losses
-      })
-
-      log(`${fileUri}:`)
-      log(`ground truth boxes: ${groundTruthBoxes.length}`)
-      log(`noObjectLoss: ${losses.noObjectLoss}`)
-      log(`objectLoss: ${losses.objectLoss}`)
-      log(`coordLoss: ${losses.coordLoss}`)
-      log(`totalLoss: ${losses.totalLoss}`)
-
-      if (Object.keys(losses).map(k => losses[k]).some(loss => isNaN(loss) || loss === Infinity)) {
-        console.log(groundTruthBoxes)
-        console.log(img)
-        console.log(losses)
-        throw new Error('corrupted loss value')
-      }
-      window.lossMap = window.lossMap || {}
-      window.lossMap[inputSize] = window.lossMap[inputSize] || {}
-      window.lossMap[inputSize][fileUri] = losses
-    }
-
-    $(document).ready(function() {
-      run()
-    })
-  </script>
-</body>
-</html>
\ No newline at end of file

tools/train/tinyYolov2/train.html

-<!DOCTYPE html>
-<html>
-<head>
-  <link rel="stylesheet" href="styles.css">
-  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
-  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
-  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
-  <script src="face-api.js"></script>
-  <script src="commons.js"></script>
-  <script src="FileSaver.js"></script>
-  <script src="trainUtils.js"></script>
-  <script src="train.js"></script>
-  <script src="loss.js"></script>
-</head>
-<body>
-
-  <script>
-    tf = faceapi.tf
-
-    const startIdx160 = 26600
-    const startIdx224 = 61660
-    const startIdx320 = 67788
-    const startIdx416 = 57650
-    const startIdx608 = 16520
-
-    //const weightsUrl = `/tmp/tmp__160_${startIdx160}__224_${startIdx224}__320_${startIdx320}__416_${startIdx416}__608_${startIdx608}.weights`
-    const weightsUrl = `/tmp/tmp_multiscale_count_8700.weights`
-
-    const fromEpoch = 0
-
-    const trainOnlyMultibox = false
-
-    const trainSizes = [416, 512, 608]
-    //const trainSizes = [608]
-
-    window.debug = false
-    window.logTrainSteps = true
-    window.count = 0
-
-
-    // hyper parameters
-    window.objectScale = 5
-    window.noObjectScale = 1
-    window.coordScale = 1
-
-    const rescaleEveryNthBatch = 100
-    window.saveEveryNthDataIdx = trainSizes.length * rescaleEveryNthBatch
-    window.trainSteps = 4000
-    //window.optimizer = tf.train.sgd(0.001)
-    window.optimizer = tf.train.adam(0.001, 0.9, 0.999, 1e-8)
-
-    // all samples
-    //const dataStartIdx = 8000
-    const dataStartIdx = 0
-    const numTrainSamples = Infinity
-
-    async function loadNetWeights(uri) {
-      return new Float32Array(await (await fetch(uri)).arrayBuffer())
-    }
-
-    async function fetchDetectionFilenames() {
-      return fetch('/detection_filenames').then(res => res.json())
-    }
-
-    async function fetchDetectionFilenamesMultibox() {
-      return fetch('/detection_filenames_multibox').then(res => res.json())
-    }
-
-    async function run() {
-      const weights = await loadNetWeights(weightsUrl)
-      window.net = new faceapi.TinyYolov2(true)
-      window.net.load(weights)
-      window.net.variable()
-
-      const fetchDetectionsFn = trainOnlyMultibox
-        ? fetchDetectionFilenamesMultibox
-        : fetchDetectionFilenames
-
-      window.detectionFilenames = (await fetchDetectionsFn()).slice(dataStartIdx, dataStartIdx + numTrainSamples)
-
-      window.lossMap = {}
-
-      console.log('ready')
-    }
-
-    function logLossChange(lossType) {
-      const { currentLoss, prevLoss, detectionFilenames } = window
-      log(`${lossType} : ${faceapi.round(currentLoss[lossType])} (avg: ${faceapi.round(currentLoss[lossType] / detectionFilenames.length)}) (delta: ${currentLoss[lossType] - prevLoss[lossType]})`)
-    }
-
-
-    function onBatchProcessed(dataIdx, inputSize) {
-      window.count++
-      const idx = (dataIdx + 1) + (window.epoch * window.detectionFilenames.length)
-      console.log('dataIdx', dataIdx)
-      if ((window.count % saveEveryNthDataIdx) === 0) {
-        saveWeights(window.net, `tmp_multiscale_count_${window.count}.weights`)
-      }
-    }
-
-    function _onBatchProcessed(dataIdx, inputSize) {
-      const idx = (dataIdx + 1) + (window.epoch * window.detectionFilenames.length)
-      console.log('idx', idx)
-      if ((idx % saveEveryNthDataIdx) === 0) {
-        saveWeights(window.net, `tmp__224_${startIdx224 + (inputSize === 224 ? idx : 0)}__320_${startIdx320 + (inputSize === 320 ? idx : 0)}__416_${startIdx416 + (inputSize === 416 ? idx : 0)}__608_${startIdx608 + (inputSize === 608 ? idx : 0)}.weights`)
-      }
-    }
-
-    async function train() {
-
-      const batchSize = 1
-
-      for (let i = fromEpoch; i < trainSteps; i++) {
-        window.epoch = i
-        log('step', i)
-        let ts2 = Date.now()
-
-        const batchCreators = createBatchCreators(shuffle(window.detectionFilenames), batchSize)
-
-        await trainStep(batchCreators, trainSizes, rescaleEveryNthBatch, onBatchProcessed)
-
-        ts2 = Date.now() - ts2
-      }
-
-      ts = Date.now() - ts
-
-      log()
-      log('--------------------')
-      log()
-      log('step %s done (%s ms)', i, ts)
-
-      window.prevLoss = window.currentLoss
-      window.currentLoss = Object.keys(lossMap)
-        .map(filename => lossMap[filename])
-        .reduce((accumulatedLosses, losses) =>
-          Object.keys(losses)
-            .map(key => ({
-              [key]: (accumulatedLosses[key] || 0) + losses[key]
-            }))
-            .reduce((map, curr) => ({ ...map, ...curr }), {}),
-          {}
-        )
-
-      if (window.prevLoss) {
-        logLossChange('noObjectLoss')
-        logLossChange('objectLoss')
-        logLossChange('coordLoss')
-        logLossChange('totalLoss')
-      }
-      log()
-      log('--------------------')
-      log()
-    }
-
-    run()
-
-  </script>
-
-</body>
-</html>
\ No newline at end of file

tools/train/tinyYolov2/train.js

-const batchIdx = 0
-
-function minimize(groundTruthBoxes, batchInput, inputSize, batch, { reshapedImgDims, paddings }) {
-  const filename = batch.filenames[batchIdx]
-  const { dataIdx } = batch
-
-  return optimizer.minimize(() => {
-
-    const outTensor = window.net.forwardInput(batchInput, inputSize)
-
-    const {
-      noObjectLoss,
-      objectLoss,
-      coordLoss,
-      totalLoss
-    } = computeLoss(
-      outTensor,
-      groundTruthBoxes,
-      reshapedImgDims,
-      paddings
-    )
-
-    const losses = {
-      totalLoss: totalLoss.dataSync()[0],
-      noObjectLoss: noObjectLoss.dataSync()[0],
-      objectLoss: objectLoss.dataSync()[0],
-      coordLoss: coordLoss.dataSync()[0]
-    }
-
-    const lossKey = `${filename}_${inputSize}`
-
-    if (window.logTrainSteps) {
-      log(`ground truth boxes: ${groundTruthBoxes.length}`)
-      log(`noObjectLoss[${dataIdx}]: ${losses.noObjectLoss}`)
-      log(`objectLoss[${dataIdx}]: ${losses.objectLoss}`)
-      log(`coordLoss[${dataIdx}]: ${losses.coordLoss}`)
-      log(`totalLoss[${dataIdx}]: ${losses.totalLoss}`)
-
-      if (window.lossMap[lossKey]) {
-        log(`loss change: ${losses.totalLoss - window.lossMap[lossKey].totalLoss}`)
-      }
-    }
-
-    window.lossMap[lossKey] = losses
-
-    return totalLoss
-  }, true)
-}
-
-async function trainStep(batchCreators, inputSizes, rescaleEveryNthBatch, onBatchProcessed = () => {}) {
-
-  async function step(currentBatchCreators) {
-    if (!currentBatchCreators.batchCreators.length) {
-      return
-    }
-
-    await promiseSequential(inputSizes.map(inputSize => async () => {
-
-      await promiseSequential(currentBatchCreators.batchCreators.map(batchCreator => async () => {
-
-        const batch = await batchCreator()
-        const { imgs, groundTruthBoxes, filenames, dataIdx } = batch
-
-        const img = imgs[0]
-        const { reshapedImgDims, paddings } = getPaddingsAndReshapedSize(img, inputSize)
-        const squareImg = imageToSquare(img)
-
-        const batchInput = await faceapi.toNetInput(squareImg)
-
-        const [imgHeight, imgWidth] = batchInput.inputs[batchIdx].shape
-
-        // skip groundTruthBoxes, which are too tiny
-        const scaleFactor = inputSize / Math.max(imgHeight, imgWidth)
-        const filteredGroundTruthBoxes = groundTruthBoxes[batchIdx].filter(({ x, y, width, height }) => {
-          const box = (new faceapi.Rect(x, y, width, height))
-            .toBoundingBox()
-            .rescale({ height: imgHeight, width: imgWidth })
-            .rescale(scaleFactor)
-
-          const isTooTiny = box.width < 40 || box.height < 40
-          if (isTooTiny && window.debug) {
-            log(`skipping box for input size ${inputSize}: (${Math.floor(box.width)} x ${Math.floor(box.height)})`)
-          }
-          return !isTooTiny
-        })
-
-        if (!filteredGroundTruthBoxes.length) {
-          if (window.debug) {
-            log(`no boxes for input size ${inputSize}, ${groundTruthBoxes[batchIdx].length} boxes were too small`)
-          }
-          batchInput.dispose()
-          onBatchProcessed(dataIdx, inputSize)
-          return
-        }
-
-        let ts = Date.now()
-        const loss = minimize(filteredGroundTruthBoxes, batchInput, inputSize, batch, { reshapedImgDims, paddings })
-
-        ts = Date.now() - ts
-        if (window.logTrainSteps) {
-          log(`trainStep time for dataIdx ${dataIdx} (${inputSize}): ${ts} ms`)
-        }
-        loss.dispose()
-        batchInput.dispose()
-        onBatchProcessed(dataIdx, inputSize)
-
-        await tf.nextFrame()
-      }))
-
-    }))
-
-    await step(currentBatchCreators.next(rescaleEveryNthBatch))
-  }
-
-  await step(batchCreators.next(rescaleEveryNthBatch))
-}
-
-async function fetchGroundTruthBoxesForFile(file) {
-  const boxes = await fetch(file).then(res => res.json())
-  return {
-    file,
-    boxes
-  }
-}
-
-function createBatchCreators(detectionFilenames, batchSize) {
-  if (batchSize < 1) {
-    throw new Error('invalid batch size: ' + batchSize)
-  }
-
-  const batches = []
-  const pushToBatch = (remaining) => {
-    if (remaining.length) {
-      batches.push(remaining.slice(0, batchSize))
-      pushToBatch(remaining.slice(batchSize))
-    }
-    return batches
-  }
-
-  pushToBatch(detectionFilenames)
-
-  const batchCreators = batches.map((filenamesForBatch, dataIdx) => async () => {
-    const groundTruthBoxes = (await Promise.all(filenamesForBatch.map(fetchGroundTruthBoxesForFile)))
-      .map(({ boxes }) => boxes)
-
-    const imgs = await Promise.all(filenamesForBatch.map(
-      async file => await faceapi.bufferToImage(await fetchImage(file.replace('.json', '')))
-    ))
-
-    return {
-      imgs,
-      groundTruthBoxes,
-      filenames: filenamesForBatch,
-      dataIdx
-    }
-  })
-
-  let idx = 0
-  function next(n) {
-    const nextBatchCreators = batchCreators.slice(idx, idx + n)
-    idx += n
-
-    return {
-      batchCreators: nextBatchCreators,
-      next
-    }
-  }
-
-  return {
-    data: batchCreators,
-    next
-  }
-}
\ No newline at end of file

tools/train/tinyYolov2/verify.html

-<!DOCTYPE html>
-<html>
-<head>
-  <script src="face-api.js"></script>
-  <script src="commons.js"></script>
-  <link rel="stylesheet" href="styles.css">
-  <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/css/materialize.css">
-  <script type="text/javascript" src="https://code.jquery.com/jquery-2.1.1.min.js"></script>
-  <script src="https://cdnjs.cloudflare.com/ajax/libs/materialize/0.100.2/js/materialize.min.js"></script>
-</head>
-<body>
-  <div id="navbar"></div>
-  <div class="center-content page-container">
-    <div class="progress" id="loader">
-      <div class="indeterminate"></div>
-    </div>
-    <div style="position: relative" class="margin">
-      <img id="inputImg" src="" style="max-width: 800px;" />
-      <canvas id="overlay" />
-    </div>
-    <div class="row side-by-side">
-      <div id="selectList"></div>
-      <div class="row">
-        <label for="imgUrlInput">Get image from URL:</label>
-        <input id="imgUrlInput" type="text" class="bold">
-      </div>
-      <button
-        class="waves-effect waves-light btn"
-        onclick="loadImageFromUrl()"
-      >
-        Ok
-      </button>
-    </div>
-    <div class="row side-by-side">
-      <div class="row input-field" style="margin-right: 20px;">
-        <select id="sizeType">
-          <option value="128">128 x 128</option>
-          <option value="160">160 x 160</option>
-          <option value="224">224 x 224</option>
-          <option value="320">320 x 320</option>
-          <option value="416">416 x 416</option>
-          <option value="608">608 x 608</option>
-        </select>
-        <label>Input Size</label>
-      </div>
-      <div class="row">
-        <label for="scoreThreshold">Score Threshold:</label>
-        <input disabled value="0.5" id="scoreThreshold" type="text" class="bold">
-      </div>
-      <button
-        class="waves-effect waves-light btn"
-        onclick="onDecreaseThreshold()"
-      >
-        <i class="material-icons left">-</i>
-      </button>
-      <button
-        class="waves-effect waves-light btn"
-        onclick="onIncreaseThreshold()"
-      >
-        <i class="material-icons left">+</i>
-      </button>
-    </div>
-    <div class="row">
-      <label for="imgByNr">Enter image NR: </label>
-      <input id="imgByNr" type="text" class="bold">
-    </div>
-  </div>
-
-  <script>
-    let scoreThreshold = 0.5
-    let sizeType = 608
-
-    function onKeyDown(e) {
-      e.target.value = (
-        parseInt(e.target.value) + (e.keyCode === 38 ? 1 : (e.keyCode === 40 ? -1 : 0))
-      ) || e.target.value || 0
-
-
-      const imgUri = window.imgs[e.target.value]
-
-      console.log(imgUri)
-      onSelectionChanged(imgUri)
-    }
-
-    function onIncreaseThreshold() {
-      scoreThreshold = Math.min(faceapi.round(scoreThreshold + 0.1), 1.0)
-      $('#scoreThreshold').val(scoreThreshold)
-      updateResults()
-    }
-
-    function onDecreaseThreshold() {
-      scoreThreshold = Math.max(faceapi.round(scoreThreshold - 0.1), 0.1)
-      $('#scoreThreshold').val(scoreThreshold)
-      updateResults()
-    }
-
-    function onSizeTypeChanged(e, c) {
-      sizeType = e.target.value
-      $('#sizeType').val(sizeType)
-      updateResults()
-    }
-
-    async function loadImageFromUrl(url) {
-      const img = await requestExternalImage($('#imgUrlInput').val())
-      $('#inputImg').get(0).src = img.src
-      updateResults()
-    }
-
-    async function updateResults() {
-      const inputImgEl = $('#inputImg').get(0)
-      const { width, height } = inputImgEl
-      const canvas = $('#overlay').get(0)
-      canvas.width = width
-      canvas.height = height
-
-      const forwardParams = {
-        inputSize: parseInt(sizeType),
-        scoreThreshold
-      }
-
-      const detections = await window.net.locateFaces(inputImgEl, forwardParams)
-
-      faceapi.drawDetection('overlay', detections.map(det => det.forSize(width, height)))
-    }
-
-    async function onSelectionChanged(uri) {
-      const imgBuf = await fetchImage(uri)
-      $(`#inputImg`).get(0).src = (await faceapi.bufferToImage(imgBuf)).src
-      updateResults()
-    }
-
-    async function loadNetWeights(uri) {
-      return new Float32Array(await (await fetch(uri)).arrayBuffer())
-    }
-
-    async function fetchDetectionFilenames() {
-      return fetch('/detection_filenames').then(res => res.json())
-    }
-
-    async function run() {
-      $('#imgByNr').keydown(onKeyDown)
-
-      const startIdx224 = 35060
-      const startIdx320 = 41188
-      const startIdx416 = 31050
-      const startIdx608 = 16520
-
-      //const weightsUrl = `/tmp/tmp__224_${startIdx224}__320_${startIdx320}__416_${startIdx416}__608_${startIdx608}.weights`
-      const weightsUrl = `/tmp/tmp_multiscale_count_4200.weights`
-
-      const weights = await loadNetWeights(weightsUrl)
-      window.net = new faceapi.TinyYolov2(true)
-      await window.net.load(weights)
-
-      window.imgs = (await fetchDetectionFilenames()).map(f => f.replace('.json', ''))
-
-      $('#loader').hide()
-      onSelectionChanged($('#selectList select').val())
-    }
-
-    $(document).ready(function() {
-      renderNavBar('#navbar', 'tiny_yolov2_face_detection')
-      renderImageSelectList(
-        '#selectList',
-        async (uri) => {
-          await onSelectionChanged(uri)
-        },
-        'bbt1.jpg'
-      )
-
-      const sizeTypeSelect = $('#sizeType')
-      sizeTypeSelect.val(sizeType)
-      sizeTypeSelect.on('change', onSizeTypeChanged)
-      sizeTypeSelect.material_select()
-      run()
-    })
-  </script>
-</body>
-</html>
\ No newline at end of file

weights/face_landmark_68_model-weights_manifest.json

-[{"weights": [{"name": "unused_Reshape/shape", "dtype": "int32", "shape": [4], "quantization": {"dtype": "uint8", "scale": 0.5058823529411764, "min": -1.011764705882353}}, {"name": "conv2d_0/kernel", "dtype": "float32", "shape": [3, 3, 3, 32], "quantization": {"dtype": "uint8", "scale": 0.006229220535240922, "min": -1.0963428142024023}}, {"name": "conv2d_0/bias", "dtype": "float32", "shape": [32], "quantization": {"dtype": "uint8", "scale": 0.009229176067838482, "min": -2.0211895588566273}}, {"name": "conv2d_1/kernel", "dtype": "float32", "shape": [3, 3, 32, 64], "quantization": {"dtype": "uint8", "scale": 0.00794055274888581, "min": -1.1513801485884423}}, {"name": "conv2d_1/bias", "dtype": "float32", "shape": [64], "quantization": {"dtype": "uint8", "scale": 0.004187015750828911, "min": -0.9839487014447941}}, {"name": "conv2d_2/kernel", "dtype": "float32", "shape": [3, 3, 64, 64], "quantization": {"dtype": "uint8", "scale": 0.013999465633841121, "min": -2.6318995391621307}}, {"name": "conv2d_2/bias", "dtype": "float32", "shape": [64], "quantization": {"dtype": "uint8", "scale": 0.0037124345699946085, "min": -0.8427226473887761}}, {"name": "conv2d_3/kernel", "dtype": "float32", "shape": [3, 3, 64, 64], "quantization": {"dtype": "uint8", "scale": 0.008715396301419128, "min": -1.368317219322803}}, {"name": "conv2d_3/bias", "dtype": "float32", "shape": [64], "quantization": {"dtype": "uint8", "scale": 0.0022804437753032236, "min": -0.5290629558703479}}, {"name": "conv2d_4/kernel", "dtype": "float32", "shape": [3, 3, 64, 64], "quantization": {"dtype": "uint8", "scale": 0.011885881891437605, "min": -2.2226599136988323}}, {"name": "conv2d_4/bias", "dtype": "float32", "shape": [64], "quantization": {"dtype": "uint8", "scale": 0.0020124191893082038, "min": -0.4809681862446607}}, {"name": "conv2d_5/kernel", "dtype": "float32", "shape": [3, 3, 64, 128], "quantization": {"dtype": "uint8", "scale": 0.011104925941018497, "min": -1.8212078543270336}}, {"name": "conv2d_5/bias", "dtype": "float32", "shape": [128], "quantization": {"dtype": "uint8", "scale": 0.0011017150592570212, "min": -0.24017388291803063}}, {"name": "conv2d_6/kernel", "dtype": "float32", "shape": [3, 3, 128, 128], "quantization": {"dtype": "uint8", "scale": 0.012181366191190831, "min": -2.009925421546487}}, {"name": "conv2d_6/bias", "dtype": "float32", "shape": [128], "quantization": {"dtype": "uint8", "scale": 0.0015788526511659809, "min": -0.3347167620471879}}, {"name": "conv2d_7/kernel", "dtype": "float32", "shape": [3, 3, 128, 256], "quantization": {"dtype": "uint8", "scale": 0.013691982100991642, "min": -2.3550209213705626}}, {"name": "conv2d_7/bias", "dtype": "float32", "shape": [256], "quantization": {"dtype": "uint8", "scale": 0.0009620819898212657, "min": -0.1452743804630111}}, {"name": "unused_flatten/Reshape/shape", "dtype": "int32", "shape": [2], "quantization": {"dtype": "uint8", "scale": 0.00784313725490196, "min": -1.003921568627451}}, {"name": "dense/kernel", "dtype": "float32", "shape": [6400, 1024], "quantization": {"dtype": "uint8", "scale": 0.01880466704275094, "min": -3.2720120654386635}}, {"name": "dense/bias", "dtype": "float32", "shape": [1024], "quantization": {"dtype": "uint8", "scale": 0.00283245981908312, "min": -0.24642400426023145}}, {"name": "logits/kernel", "dtype": "float32", "shape": [1024, 136], "quantization": {"dtype": "uint8", "scale": 0.005545294986051672, "min": -0.7818865930332858}}, {"name": "logits/bias", "dtype": "float32", "shape": [136], "quantization": {"dtype": "uint8", "scale": 0.002829533581640206, "min": 0.13660642504692078}}], "paths": ["face_landmark_68_model-shard1", "face_landmark_68_model-shard2"]}]
\ No newline at end of file
+[{"weights":[{"name":"dense0/conv0/filters","shape":[3,3,3,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004853619781194949,"min":-0.5872879935245888}},{"name":"dense0/conv0/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004396426443960153,"min":-0.7298067896973853}},{"name":"dense0/conv1/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00635151559231328,"min":-0.5589333721235686}},{"name":"dense0/conv1/pointwise_filter","shape":[1,1,32,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.009354315552057004,"min":-1.2628325995276957}},{"name":"dense0/conv1/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0029380727048013726,"min":-0.5846764682554731}},{"name":"dense0/conv2/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0049374802439820535,"min":-0.6171850304977566}},{"name":"dense0/conv2/pointwise_filter","shape":[1,1,32,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.009941946758943446,"min":-1.3421628124573652}},{"name":"dense0/conv2/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0030300481062309416,"min":-0.5272283704841838}},{"name":"dense0/conv3/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.005672684837790097,"min":-0.7431217137505026}},{"name":"dense0/conv3/pointwise_filter","shape":[1,1,32,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.010712201455060173,"min":-1.5639814124387852}},{"name":"dense0/conv3/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0030966934035806097,"min":-0.3839899820439956}},{"name":"dense1/conv0/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0039155554537679636,"min":-0.48161332081345953}},{"name":"dense1/conv0/pointwise_filter","shape":[1,1,32,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.01023082966898002,"min":-1.094698774580862}},{"name":"dense1/conv0/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0027264176630506327,"min":-0.3871513081531898}},{"name":"dense1/conv1/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004583378632863362,"min":-0.5454220573107401}},{"name":"dense1/conv1/pointwise_filter","shape":[1,1,64,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00915846403907327,"min":-1.117332612766939}},{"name":"dense1/conv1/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003091680419211294,"min":-0.5966943209077797}},{"name":"dense1/conv2/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.005407439727409214,"min":-0.708374604290607}},{"name":"dense1/conv2/pointwise_filter","shape":[1,1,64,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00946493943532308,"min":-1.2399070660273235}},{"name":"dense1/conv2/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004409168514550901,"min":-0.9788354102303}},{"name":"dense1/conv3/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004478132958505668,"min":-0.6493292789833219}},{"name":"dense1/conv3/pointwise_filter","shape":[1,1,64,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.011063695888893277,"min":-1.2501976354449402}},{"name":"dense1/conv3/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003909627596537272,"min":-0.6646366914113363}},{"name":"dense2/conv0/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003213915404151468,"min":-0.3374611174359041}},{"name":"dense2/conv0/pointwise_filter","shape":[1,1,64,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.010917326048308728,"min":-1.4520043644250609}},{"name":"dense2/conv0/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.002800439152063108,"min":-0.38085972468058266}},{"name":"dense2/conv1/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0050568851770139206,"min":-0.6927932692509071}},{"name":"dense2/conv1/pointwise_filter","shape":[1,1,128,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.01074961213504567,"min":-1.3222022926106174}},{"name":"dense2/conv1/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0030654204242369708,"min":-0.5487102559384177}},{"name":"dense2/conv2/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00591809165244009,"min":-0.917304206128214}},{"name":"dense2/conv2/pointwise_filter","shape":[1,1,128,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.01092823346455892,"min":-1.366029183069865}},{"name":"dense2/conv2/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.002681120470458386,"min":-0.36463238398234055}},{"name":"dense2/conv3/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0048311497650894465,"min":-0.5797379718107336}},{"name":"dense2/conv3/pointwise_filter","shape":[1,1,128,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.011227761062921263,"min":-1.4483811771168429}},{"name":"dense2/conv3/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0034643323982463162,"min":-0.3360402426298927}},{"name":"dense3/conv0/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003394978887894574,"min":-0.49227193874471326}},{"name":"dense3/conv0/pointwise_filter","shape":[1,1,128,256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.010051267287310432,"min":-1.2765109454884247}},{"name":"dense3/conv0/bias","shape":[256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003142924752889895,"min":-0.4588670139219247}},{"name":"dense3/conv1/depthwise_filter","shape":[3,3,256,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00448304671867221,"min":-0.5872791201460595}},{"name":"dense3/conv1/pointwise_filter","shape":[1,1,256,256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.016063522357566685,"min":-2.3613377865623026}},{"name":"dense3/conv1/bias","shape":[256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00287135781026354,"min":-0.47664539650374765}},{"name":"dense3/conv2/depthwise_filter","shape":[3,3,256,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.006002906724518421,"min":-0.7923836876364315}},{"name":"dense3/conv2/pointwise_filter","shape":[1,1,256,256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.017087187019048954,"min":-1.6061955797906016}},{"name":"dense3/conv2/bias","shape":[256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.003124481205846749,"min":-0.46242321846531886}},{"name":"dense3/conv3/depthwise_filter","shape":[3,3,256,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.006576311588287353,"min":-1.0193282961845398}},{"name":"dense3/conv3/pointwise_filter","shape":[1,1,256,256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.015590153955945782,"min":-1.99553970636106}},{"name":"dense3/conv3/bias","shape":[256],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.004453541601405424,"min":-0.6546706154065973}},{"name":"fc/weights","shape":[256,136],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.010417488509533453,"min":-1.500118345372817}},{"name":"fc/bias","shape":[136],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0025084222648658005,"min":0.07683877646923065}}],"paths":["face_landmark_68_model-shard1"]}]
\ No newline at end of file

weights/face_landmark_68_tiny_model-weights_manifest.json

+[{"weights":[{"name":"dense0/conv0/filters","shape":[3,3,3,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.008194216092427571,"min":-0.9423348506291708}},{"name":"dense0/conv0/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.006839508168837603,"min":-0.8412595047670252}},{"name":"dense0/conv1/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.009194007106855804,"min":-1.2779669878529567}},{"name":"dense0/conv1/pointwise_filter","shape":[1,1,32,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0036026100317637128,"min":-0.3170296827952067}},{"name":"dense0/conv1/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.000740380117706224,"min":-0.06367269012273527}},{"name":"dense0/conv2/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":1,"min":0}},{"name":"dense0/conv2/pointwise_filter","shape":[1,1,32,32],"dtype":"float32","quantization":{"dtype":"uint8","scale":1,"min":0}},{"name":"dense0/conv2/bias","shape":[32],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0037702228508743585,"min":-0.6220867703942692}},{"name":"dense1/conv0/depthwise_filter","shape":[3,3,32,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0033707996209462483,"min":-0.421349952618281}},{"name":"dense1/conv0/pointwise_filter","shape":[1,1,32,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.014611541991140328,"min":-1.8556658328748217}},{"name":"dense1/conv0/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.002832523046755323,"min":-0.30307996600281956}},{"name":"dense1/conv1/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.006593170586754294,"min":-0.6329443763284123}},{"name":"dense1/conv1/pointwise_filter","shape":[1,1,64,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.012215249211180444,"min":-1.6001976466646382}},{"name":"dense1/conv1/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.002384825547536214,"min":-0.3028728445370992}},{"name":"dense1/conv2/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.005859645441466687,"min":-0.7617539073906693}},{"name":"dense1/conv2/pointwise_filter","shape":[1,1,64,64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.013121426806730382,"min":-1.7845140457153321}},{"name":"dense1/conv2/bias","shape":[64],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0032247188044529336,"min":-0.46435950784122243}},{"name":"dense2/conv0/depthwise_filter","shape":[3,3,64,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.002659512618008782,"min":-0.32977956463308894}},{"name":"dense2/conv0/pointwise_filter","shape":[1,1,64,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.015499923743453681,"min":-1.9839902391620712}},{"name":"dense2/conv0/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0032450980999890497,"min":-0.522460794098237}},{"name":"dense2/conv1/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.005911862382701799,"min":-0.792189559282041}},{"name":"dense2/conv1/pointwise_filter","shape":[1,1,128,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.021025861478319356,"min":-2.2077154552235325}},{"name":"dense2/conv1/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.00349616945958605,"min":-0.46149436866535865}},{"name":"dense2/conv2/depthwise_filter","shape":[3,3,128,1],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.008104994250278847,"min":-1.013124281284856}},{"name":"dense2/conv2/pointwise_filter","shape":[1,1,128,128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.029337059282789044,"min":-3.5791212325002633}},{"name":"dense2/conv2/bias","shape":[128],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0038808938334969913,"min":-0.4230174278511721}},{"name":"fc/weights","shape":[128,136],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.014016061670639936,"min":-1.8921683255363912}},{"name":"fc/bias","shape":[136],"dtype":"float32","quantization":{"dtype":"uint8","scale":0.0029505149698724935,"min":0.088760145008564}}],"paths":["face_landmark_68_tiny_model-shard1"]}]
\ No newline at end of file