mirror of https://github.com/vladmandic/human
pure tensor pipeline without image converts
Vladimir Mandic
parent
5e90ab0d51
commit
b146a0a64e
|
|
@ -268,7 +268,6 @@ config = {
|
|||
inputSize: 64, // fixed value
|
||||
minConfidence: 0.5, // threshold for discarding a prediction
|
||||
skipFrames: 10, // how many frames to go without re-running the detector, only used for video inputs
|
||||
useGrayscale: true, // convert image to grayscale before prediction or use highest channel
|
||||
modelPath: '../models/emotion/model.json',
|
||||
},
|
||||
},
|
||||
|
|
|
|||
|
|
@ -51,7 +51,6 @@ export default {
|
|||
inputSize: 64, // fixed value
|
||||
minConfidence: 0.5, // threshold for discarding a prediction
|
||||
skipFrames: 10, // how many frames to go without re-running the detector
|
||||
useGrayscale: true, // convert image to grayscale before prediction or use highest channel
|
||||
modelPath: '../models/emotion/model.json',
|
||||
},
|
||||
},
|
||||
|
|
|
|||
|
|
@ -144,13 +144,7 @@ var require_blazeface = __commonJS((exports2) => {
|
|||
};
|
||||
}
|
||||
async estimateFaces(input) {
|
||||
const imageRaw = !(input instanceof tf2.Tensor) ? tf2.browser.fromPixels(input) : input;
|
||||
const imageCast = imageRaw.toFloat();
|
||||
const image = imageCast.expandDims(0);
|
||||
imageRaw.dispose();
|
||||
imageCast.dispose();
|
||||
const {boxes, scaleFactor} = await this.getBoundingBoxes(image);
|
||||
image.dispose();
|
||||
const {boxes, scaleFactor} = await this.getBoundingBoxes(input);
|
||||
return Promise.all(boxes.map(async (face) => {
|
||||
const scaledBox = scaleBoxFromPrediction(face, scaleFactor);
|
||||
const [landmarkData, boxData, probabilityData] = await Promise.all([face.landmarks, scaledBox, face.probability].map(async (d) => d.array()));
|
||||
|
|
@ -3818,13 +3812,7 @@ var require_facemesh = __commonJS((exports2) => {
|
|||
async estimateFaces(input, config2) {
|
||||
if (config2)
|
||||
this.config = config2;
|
||||
const imageRaw = !(input instanceof tf2.Tensor) ? tf2.browser.fromPixels(input) : input;
|
||||
const imageCast = imageRaw.toFloat();
|
||||
const image = imageCast.expandDims(0);
|
||||
imageRaw.dispose();
|
||||
imageCast.dispose();
|
||||
const predictions = await this.pipeline.predict(image, config2);
|
||||
tf2.dispose(image);
|
||||
const predictions = await this.pipeline.predict(input, config2);
|
||||
const results = [];
|
||||
for (const prediction of predictions || []) {
|
||||
if (prediction.isDisposedInternal)
|
||||
|
|
@ -3879,12 +3867,6 @@ var require_ssrnet = __commonJS((exports2) => {
|
|||
const models2 = {};
|
||||
let last = {age: 0, gender: ""};
|
||||
let frame = 0;
|
||||
async function getImage(image, size) {
|
||||
const buffer = tf2.browser.fromPixels(image);
|
||||
const resize = tf2.image.resizeBilinear(buffer, [size, size]);
|
||||
const expand = tf2.cast(tf2.expandDims(resize, 0), "float32");
|
||||
return expand;
|
||||
}
|
||||
async function loadAge(config2) {
|
||||
if (!models2.age)
|
||||
models2.age = await tf2.loadGraphModel(config2.face.age.modelPath);
|
||||
|
|
@ -3901,14 +3883,9 @@ var require_ssrnet = __commonJS((exports2) => {
|
|||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
let enhance;
|
||||
if (image instanceof tf2.Tensor) {
|
||||
const resize = tf2.image.resizeBilinear(image, [config2.face.age.inputSize, config2.face.age.inputSize], false);
|
||||
enhance = tf2.mul(resize, [255]);
|
||||
tf2.dispose(resize);
|
||||
} else {
|
||||
enhance = await getImage(image, config2.face.age.inputSize);
|
||||
}
|
||||
const resize = tf2.image.resizeBilinear(image, [config2.face.age.inputSize, config2.face.age.inputSize], false);
|
||||
const enhance = tf2.mul(resize, [255]);
|
||||
tf2.dispose(resize);
|
||||
const promises = [];
|
||||
let ageT;
|
||||
let genderT;
|
||||
|
|
@ -3949,15 +3926,6 @@ var require_emotion = __commonJS((exports2) => {
|
|||
let last = [];
|
||||
let frame = 0;
|
||||
const multiplier = 1.5;
|
||||
function getImage(image, size) {
|
||||
const tensor = tf2.tidy(() => {
|
||||
const buffer = tf2.browser.fromPixels(image, 1);
|
||||
const resize = tf2.image.resizeBilinear(buffer, [size, size]);
|
||||
const expand = tf2.cast(tf2.expandDims(resize, 0), "float32");
|
||||
return expand;
|
||||
});
|
||||
return tensor;
|
||||
}
|
||||
async function load2(config2) {
|
||||
if (!models2.emotion)
|
||||
models2.emotion = await tf2.loadGraphModel(config2.face.emotion.modelPath);
|
||||
|
|
@ -3969,24 +3937,22 @@ var require_emotion = __commonJS((exports2) => {
|
|||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
const enhance = tf2.tidy(() => {
|
||||
if (image instanceof tf2.Tensor) {
|
||||
const resize = tf2.image.resizeBilinear(image, [config2.face.emotion.inputSize, config2.face.emotion.inputSize], false);
|
||||
const [r, g, b] = tf2.split(resize, 3, 3);
|
||||
if (config2.face.emotion.useGrayscale) {
|
||||
const r1 = tf2.mul(r, [0.2989]);
|
||||
const g1 = tf2.mul(g, [0.587]);
|
||||
const b1 = tf2.mul(b, [0.114]);
|
||||
const grayscale = tf2.addN([r1, g1, b1]);
|
||||
return grayscale;
|
||||
}
|
||||
return g;
|
||||
}
|
||||
return getImage(image, config2.face.emotion.inputSize);
|
||||
});
|
||||
const resize = tf2.image.resizeBilinear(image, [config2.face.emotion.inputSize, config2.face.emotion.inputSize], false);
|
||||
const [red, green, blue] = tf2.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf2.mul(red, [0.2989]);
|
||||
const greenNorm = tf2.mul(green, [0.587]);
|
||||
const blueNorm = tf2.mul(blue, [0.114]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf2.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const obj = [];
|
||||
if (config2.face.emotion.enabled) {
|
||||
const emotionT = await models2.emotion.predict(enhance);
|
||||
const emotionT = await models2.emotion.predict(grayscale);
|
||||
const data = await emotionT.data();
|
||||
for (let i = 0; i < data.length; i++) {
|
||||
if (multiplier * data[i] > config2.face.emotion.minConfidence)
|
||||
|
|
@ -3995,7 +3961,7 @@ var require_emotion = __commonJS((exports2) => {
|
|||
obj.sort((a, b) => b.score - a.score);
|
||||
tf2.dispose(emotionT);
|
||||
}
|
||||
tf2.dispose(enhance);
|
||||
tf2.dispose(grayscale);
|
||||
last = obj;
|
||||
return obj;
|
||||
}
|
||||
|
|
@ -4407,7 +4373,6 @@ var require_decodeMultiple = __commonJS((exports2) => {
|
|||
|
||||
// src/posenet/util.js
|
||||
var require_util2 = __commonJS((exports2) => {
|
||||
const tf2 = require("@tensorflow/tfjs");
|
||||
const kpt = require_keypoints2();
|
||||
function eitherPointDoesntMeetConfidence(a, b, minConfidence) {
|
||||
return a < minConfidence || b < minConfidence;
|
||||
|
|
@ -4446,70 +4411,26 @@ var require_util2 = __commonJS((exports2) => {
|
|||
return Promise.all(tensors.map((tensor) => tensor.buffer()));
|
||||
}
|
||||
exports2.toTensorBuffers3D = toTensorBuffers3D;
|
||||
function scalePose(pose, scaleY, scaleX, offsetY = 0, offsetX = 0) {
|
||||
function scalePose(pose, scaleY, scaleX) {
|
||||
return {
|
||||
score: pose.score,
|
||||
keypoints: pose.keypoints.map(({score, part, position}) => ({
|
||||
score,
|
||||
part,
|
||||
position: {
|
||||
x: position.x * scaleX + offsetX,
|
||||
y: position.y * scaleY + offsetY
|
||||
}
|
||||
position: {x: position.x * scaleX, y: position.y * scaleY}
|
||||
}))
|
||||
};
|
||||
}
|
||||
exports2.scalePose = scalePose;
|
||||
function scalePoses(poses, scaleY, scaleX, offsetY = 0, offsetX = 0) {
|
||||
if (scaleX === 1 && scaleY === 1 && offsetY === 0 && offsetX === 0) {
|
||||
return poses;
|
||||
}
|
||||
return poses.map((pose) => scalePose(pose, scaleY, scaleX, offsetY, offsetX));
|
||||
function resizeTo(image, [targetH, targetW]) {
|
||||
const input = image.squeeze(0);
|
||||
const resized = input.resizeBilinear([targetH, targetW]);
|
||||
input.dispose();
|
||||
return resized;
|
||||
}
|
||||
exports2.scalePoses = scalePoses;
|
||||
function getInputTensorDimensions(input) {
|
||||
return input instanceof tf2.Tensor ? [input.shape[0], input.shape[1]] : [input.height, input.width];
|
||||
}
|
||||
exports2.getInputTensorDimensions = getInputTensorDimensions;
|
||||
function toInputTensor(input) {
|
||||
return input instanceof tf2.Tensor ? input : tf2.browser.fromPixels(input);
|
||||
}
|
||||
exports2.toInputTensor = toInputTensor;
|
||||
function toResizedInputTensor(input, resizeHeight, resizeWidth) {
|
||||
return tf2.tidy(() => {
|
||||
const imageTensor = toInputTensor(input);
|
||||
return imageTensor.resizeBilinear([resizeHeight, resizeWidth]);
|
||||
});
|
||||
}
|
||||
exports2.toResizedInputTensor = toResizedInputTensor;
|
||||
function padAndResizeTo(input, [targetH, targetW]) {
|
||||
const [height, width] = getInputTensorDimensions(input);
|
||||
const targetAspect = targetW / targetH;
|
||||
const aspect = width / height;
|
||||
let [padT, padB, padL, padR] = [0, 0, 0, 0];
|
||||
if (aspect < targetAspect) {
|
||||
padT = 0;
|
||||
padB = 0;
|
||||
padL = Math.round(0.5 * (targetAspect * height - width));
|
||||
padR = Math.round(0.5 * (targetAspect * height - width));
|
||||
} else {
|
||||
padT = Math.round(0.5 * (1 / targetAspect * width - height));
|
||||
padB = Math.round(0.5 * (1 / targetAspect * width - height));
|
||||
padL = 0;
|
||||
padR = 0;
|
||||
}
|
||||
const resized = tf2.tidy(() => {
|
||||
let imageTensor = toInputTensor(input);
|
||||
imageTensor = tf2.pad3d(imageTensor, [[padT, padB], [padL, padR], [0, 0]]);
|
||||
return imageTensor.resizeBilinear([targetH, targetW]);
|
||||
});
|
||||
return {resized, padding: {top: padT, left: padL, right: padR, bottom: padB}};
|
||||
}
|
||||
exports2.padAndResizeTo = padAndResizeTo;
|
||||
function scaleAndFlipPoses(poses, [height, width], [inputResolutionHeight, inputResolutionWidth], padding) {
|
||||
const scaleY = (height + padding.top + padding.bottom) / inputResolutionHeight;
|
||||
const scaleX = (width + padding.left + padding.right) / inputResolutionWidth;
|
||||
const scaledPoses = scalePoses(poses, scaleY, scaleX, -padding.top, -padding.left);
|
||||
exports2.resizeTo = resizeTo;
|
||||
function scaleAndFlipPoses(poses, [height, width], [inputResolutionHeight, inputResolutionWidth]) {
|
||||
const scaledPoses = poses.map((pose) => scalePose(pose, height / inputResolutionHeight, width / inputResolutionWidth));
|
||||
return scaledPoses;
|
||||
}
|
||||
exports2.scaleAndFlipPoses = scaleAndFlipPoses;
|
||||
|
|
@ -4527,8 +4448,9 @@ var require_modelPoseNet = __commonJS((exports2) => {
|
|||
}
|
||||
async estimatePoses(input, config2) {
|
||||
const outputStride = config2.outputStride;
|
||||
const [height, width] = util.getInputTensorDimensions(input);
|
||||
const {resized, padding} = util.padAndResizeTo(input, [config2.inputResolution, config2.inputResolution]);
|
||||
const height = input.shape[1];
|
||||
const width = input.shape[2];
|
||||
const resized = util.resizeTo(input, [config2.inputResolution, config2.inputResolution]);
|
||||
const {heatmapScores, offsets, displacementFwd, displacementBwd} = this.baseModel.predict(resized);
|
||||
const allTensorBuffers = await util.toTensorBuffers3D([heatmapScores, offsets, displacementFwd, displacementBwd]);
|
||||
const scoresBuffer = allTensorBuffers[0];
|
||||
|
|
@ -4536,7 +4458,7 @@ var require_modelPoseNet = __commonJS((exports2) => {
|
|||
const displacementsFwdBuffer = allTensorBuffers[2];
|
||||
const displacementsBwdBuffer = allTensorBuffers[3];
|
||||
const poses = await decodeMultiple.decodeMultiplePoses(scoresBuffer, offsetsBuffer, displacementsFwdBuffer, displacementsBwdBuffer, outputStride, config2.maxDetections, config2.scoreThreshold, config2.nmsRadius);
|
||||
const resultPoses = util.scaleAndFlipPoses(poses, [height, width], [config2.inputResolution, config2.inputResolution], padding);
|
||||
const resultPoses = util.scaleAndFlipPoses(poses, [height, width], [config2.inputResolution, config2.inputResolution]);
|
||||
heatmapScores.dispose();
|
||||
offsets.dispose();
|
||||
displacementFwd.dispose();
|
||||
|
|
@ -4685,15 +4607,15 @@ var require_handdetector = __commonJS((exports2) => {
|
|||
});
|
||||
}
|
||||
async getBoundingBoxes(input) {
|
||||
const normalizedInput = tf2.tidy(() => tf2.mul(tf2.sub(input, 0.5), 2));
|
||||
const batchedPrediction = this.model.predict(normalizedInput);
|
||||
const batchedPrediction = this.model.predict(input);
|
||||
const prediction = batchedPrediction.squeeze();
|
||||
console.log(prediction);
|
||||
const scores = tf2.tidy(() => tf2.sigmoid(tf2.slice(prediction, [0, 0], [-1, 1])).squeeze());
|
||||
const rawBoxes = tf2.slice(prediction, [0, 1], [-1, 4]);
|
||||
const boxes = this.normalizeBoxes(rawBoxes);
|
||||
const boxesWithHandsTensor = await tf2.image.nonMaxSuppressionAsync(boxes, scores, this.maxHands, this.iouThreshold, this.scoreThreshold);
|
||||
const boxesWithHands = await boxesWithHandsTensor.array();
|
||||
const toDispose = [normalizedInput, batchedPrediction, boxesWithHandsTensor, prediction, boxes, rawBoxes, scores];
|
||||
const toDispose = [batchedPrediction, boxesWithHandsTensor, prediction, boxes, rawBoxes, scores];
|
||||
const detectedHands = tf2.tidy(() => {
|
||||
const detectedBoxes = [];
|
||||
for (const i in boxesWithHands) {
|
||||
|
|
@ -4709,12 +4631,16 @@ var require_handdetector = __commonJS((exports2) => {
|
|||
return detectedHands;
|
||||
}
|
||||
async estimateHandBounds(input, config2) {
|
||||
const inputHeight = input.shape[1];
|
||||
const inputWidth = input.shape[2];
|
||||
this.iouThreshold = config2.iouThreshold;
|
||||
this.scoreThreshold = config2.scoreThreshold;
|
||||
this.maxHands = config2.maxHands;
|
||||
const image = tf2.tidy(() => input.resizeBilinear([this.width, this.height]).div(255));
|
||||
const resized = input.resizeBilinear([this.width, this.height]);
|
||||
const divided = resized.div(255);
|
||||
const normalized = divided.sub(0.5);
|
||||
const image = normalized.mul(2);
|
||||
resized.dispose();
|
||||
divided.dispose();
|
||||
normalized.dispose();
|
||||
const predictions = await this.getBoundingBoxes(image);
|
||||
image.dispose();
|
||||
if (!predictions || predictions.length === 0)
|
||||
|
|
@ -4728,7 +4654,7 @@ var require_handdetector = __commonJS((exports2) => {
|
|||
const palmLandmarks = await prediction.palmLandmarks.array();
|
||||
prediction.boxes.dispose();
|
||||
prediction.palmLandmarks.dispose();
|
||||
hands.push(bounding.scaleBoxCoordinates({startPoint, endPoint, palmLandmarks}, [inputWidth / this.width, inputHeight / this.height]));
|
||||
hands.push(bounding.scaleBoxCoordinates({startPoint, endPoint, palmLandmarks}, [input.shape[2] / this.width, input.shape[1] / this.height]));
|
||||
}
|
||||
return hands;
|
||||
}
|
||||
|
|
@ -4886,7 +4812,7 @@ var require_pipeline2 = __commonJS((exports2) => {
|
|||
]);
|
||||
}
|
||||
async estimateHands(image, config2) {
|
||||
this.maxContinuousChecks = config2.skipFrames;
|
||||
this.skipFrames = config2.skipFrames;
|
||||
this.detectionConfidence = config2.minConfidence;
|
||||
this.maxHands = config2.maxHands;
|
||||
this.runsWithoutHandDetector++;
|
||||
|
|
@ -4996,14 +4922,7 @@ var require_handpose = __commonJS((exports2) => {
|
|||
this.skipFrames = config2.skipFrames;
|
||||
this.detectionConfidence = config2.minConfidence;
|
||||
this.maxHands = config2.maxHands;
|
||||
const image = tf2.tidy(() => {
|
||||
if (!(input instanceof tf2.Tensor)) {
|
||||
input = tf2.browser.fromPixels(input);
|
||||
}
|
||||
return input.toFloat().expandDims(0);
|
||||
});
|
||||
const predictions = await this.pipeline.estimateHands(image, config2);
|
||||
image.dispose();
|
||||
const predictions = await this.pipeline.estimateHands(input, config2);
|
||||
const hands = [];
|
||||
if (!predictions)
|
||||
return hands;
|
||||
|
|
@ -5094,7 +5013,6 @@ var require_config = __commonJS((exports2) => {
|
|||
inputSize: 64,
|
||||
minConfidence: 0.5,
|
||||
skipFrames: 10,
|
||||
useGrayscale: true,
|
||||
modelPath: "../models/emotion/model.json"
|
||||
}
|
||||
},
|
||||
|
|
@ -5258,7 +5176,7 @@ function mergeDeep(...objects) {
|
|||
function sanity(input) {
|
||||
if (!input)
|
||||
return "input is not defined";
|
||||
if (tf.ENV.flags.IS_BROWSER && (input instanceof ImageData || input instanceof HTMLImageElement || input instanceof HTMLCanvasElement || input instanceof HTMLVideoElement || input instanceof HTMLMediaElement)) {
|
||||
if (!(input instanceof tf.Tensor) || tf.ENV.flags.IS_BROWSER && (input instanceof ImageData || input instanceof HTMLImageElement || input instanceof HTMLCanvasElement || input instanceof HTMLVideoElement || input instanceof HTMLMediaElement)) {
|
||||
const width = input.naturalWidth || input.videoWidth || input.width || input.shape && input.shape[1] > 0;
|
||||
if (!width || width === 0)
|
||||
return "input is empty";
|
||||
|
|
@ -5293,6 +5211,19 @@ async function load(userConfig) {
|
|||
if (config.face.enabled && config.face.emotion.enabled && !models.emotion)
|
||||
models.emotion = await emotion.load(config);
|
||||
}
|
||||
function tfImage(input) {
|
||||
let image;
|
||||
if (input instanceof tf.Tensor) {
|
||||
image = tf.clone(input);
|
||||
} else {
|
||||
const pixels = tf.browser.fromPixels(input);
|
||||
const casted = pixels.toFloat();
|
||||
image = casted.expandDims(0);
|
||||
pixels.dispose();
|
||||
casted.dispose();
|
||||
}
|
||||
return image;
|
||||
}
|
||||
async function detect(input, userConfig = {}) {
|
||||
state = "config";
|
||||
const perf = {};
|
||||
|
|
@ -5332,16 +5263,17 @@ async function detect(input, userConfig = {}) {
|
|||
if (config.scoped)
|
||||
tf.engine().startScope();
|
||||
analyze("Start Detect:");
|
||||
const imageTensor = tfImage(input);
|
||||
state = "run:body";
|
||||
timeStamp = now();
|
||||
analyze("Start PoseNet");
|
||||
const poseRes = config.body.enabled ? await models.posenet.estimatePoses(input, config.body) : [];
|
||||
const poseRes = config.body.enabled ? await models.posenet.estimatePoses(imageTensor, config.body) : [];
|
||||
analyze("End PoseNet:");
|
||||
perf.body = Math.trunc(now() - timeStamp);
|
||||
state = "run:hand";
|
||||
timeStamp = now();
|
||||
analyze("Start HandPose:");
|
||||
const handRes = config.hand.enabled ? await models.handpose.estimateHands(input, config.hand) : [];
|
||||
const handRes = config.hand.enabled ? await models.handpose.estimateHands(imageTensor, config.hand) : [];
|
||||
analyze("End HandPose:");
|
||||
perf.hand = Math.trunc(now() - timeStamp);
|
||||
const faceRes = [];
|
||||
|
|
@ -5349,7 +5281,7 @@ async function detect(input, userConfig = {}) {
|
|||
state = "run:face";
|
||||
timeStamp = now();
|
||||
analyze("Start FaceMesh:");
|
||||
const faces = await models.facemesh.estimateFaces(input, config.face);
|
||||
const faces = await models.facemesh.estimateFaces(imageTensor, config.face);
|
||||
perf.face = Math.trunc(now() - timeStamp);
|
||||
for (const face of faces) {
|
||||
if (!face.image || face.image.isDisposedInternal) {
|
||||
|
|
@ -5377,9 +5309,10 @@ async function detect(input, userConfig = {}) {
|
|||
emotion: emotionData,
|
||||
iris: iris !== 0 ? Math.trunc(100 * 11.7 / iris) / 100 : 0
|
||||
});
|
||||
analyze("End FaceMesh:");
|
||||
}
|
||||
analyze("End FaceMesh:");
|
||||
}
|
||||
imageTensor.dispose();
|
||||
state = "idle";
|
||||
if (config.scoped)
|
||||
tf.engine().endScope();
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 4862,
|
||||
"bytes": 4762,
|
||||
"imports": []
|
||||
},
|
||||
"package.json": {
|
||||
|
|
@ -9,11 +9,11 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2019,
|
||||
"bytes": 1646,
|
||||
"imports": []
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytes": 7407,
|
||||
"bytes": 7161,
|
||||
"imports": []
|
||||
},
|
||||
"src/facemesh/box.js": {
|
||||
|
|
@ -21,7 +21,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytes": 2816,
|
||||
"bytes": 2568,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/facemesh/blazeface.js"
|
||||
|
|
@ -75,7 +75,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytes": 4296,
|
||||
"bytes": 4277,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/box.js"
|
||||
|
|
@ -83,7 +83,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytes": 2356,
|
||||
"bytes": 2152,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/handdetector.js"
|
||||
|
|
@ -101,7 +101,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytes": 8178,
|
||||
"bytes": 8169,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/box.js"
|
||||
|
|
@ -116,7 +116,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 8299,
|
||||
"bytes": 8784,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/facemesh/facemesh.js"
|
||||
|
|
@ -195,7 +195,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytes": 3472,
|
||||
"bytes": 3447,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/posenet/modelMobileNet.js"
|
||||
|
|
@ -229,7 +229,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytes": 4202,
|
||||
"bytes": 2260,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/posenet/keypoints.js"
|
||||
|
|
@ -245,7 +245,7 @@
|
|||
]
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytes": 1937,
|
||||
"bytes": 1574,
|
||||
"imports": []
|
||||
}
|
||||
},
|
||||
|
|
@ -253,13 +253,13 @@
|
|||
"dist/human.cjs.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 220934
|
||||
"bytes": 215962
|
||||
},
|
||||
"dist/human.cjs": {
|
||||
"imports": [],
|
||||
"inputs": {
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytesInOutput": 7398
|
||||
"bytesInOutput": 7138
|
||||
},
|
||||
"src/facemesh/keypoints.js": {
|
||||
"bytesInOutput": 2771
|
||||
|
|
@ -280,13 +280,13 @@
|
|||
"bytesInOutput": 23311
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytesInOutput": 2950
|
||||
"bytesInOutput": 2687
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytesInOutput": 2158
|
||||
"bytesInOutput": 1768
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 2133
|
||||
"bytesInOutput": 1736
|
||||
},
|
||||
"src/posenet/modelBase.js": {
|
||||
"bytesInOutput": 1120
|
||||
|
|
@ -313,10 +313,10 @@
|
|||
"bytesInOutput": 1992
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytesInOutput": 4383
|
||||
"bytesInOutput": 2404
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytesInOutput": 1976
|
||||
"bytesInOutput": 1955
|
||||
},
|
||||
"src/posenet/posenet.js": {
|
||||
"bytesInOutput": 917
|
||||
|
|
@ -325,7 +325,7 @@
|
|||
"bytesInOutput": 2813
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytesInOutput": 4135
|
||||
"bytesInOutput": 4161
|
||||
},
|
||||
"src/handpose/keypoints.js": {
|
||||
"bytesInOutput": 265
|
||||
|
|
@ -334,22 +334,22 @@
|
|||
"bytesInOutput": 2671
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytesInOutput": 7625
|
||||
"bytesInOutput": 7616
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytesInOutput": 2509
|
||||
"bytesInOutput": 2288
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 1872
|
||||
"bytesInOutput": 1844
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 2778
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 7273
|
||||
"bytesInOutput": 7694
|
||||
}
|
||||
},
|
||||
"bytes": 134728
|
||||
"bytes": 131607
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 4862,
|
||||
"bytes": 4762,
|
||||
"imports": []
|
||||
},
|
||||
"package.json": {
|
||||
|
|
@ -9,11 +9,11 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2019,
|
||||
"bytes": 1646,
|
||||
"imports": []
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytes": 7407,
|
||||
"bytes": 7161,
|
||||
"imports": []
|
||||
},
|
||||
"src/facemesh/box.js": {
|
||||
|
|
@ -21,7 +21,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytes": 2816,
|
||||
"bytes": 2568,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/facemesh/blazeface.js"
|
||||
|
|
@ -75,7 +75,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytes": 4296,
|
||||
"bytes": 4277,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/box.js"
|
||||
|
|
@ -83,7 +83,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytes": 2356,
|
||||
"bytes": 2152,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/handdetector.js"
|
||||
|
|
@ -101,7 +101,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytes": 8178,
|
||||
"bytes": 8169,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/handpose/box.js"
|
||||
|
|
@ -116,7 +116,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 8299,
|
||||
"bytes": 8784,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/facemesh/facemesh.js"
|
||||
|
|
@ -195,7 +195,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytes": 3472,
|
||||
"bytes": 3447,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/posenet/modelMobileNet.js"
|
||||
|
|
@ -229,7 +229,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytes": 4202,
|
||||
"bytes": 2260,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/posenet/keypoints.js"
|
||||
|
|
@ -245,7 +245,7 @@
|
|||
]
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytes": 1937,
|
||||
"bytes": 1574,
|
||||
"imports": []
|
||||
}
|
||||
},
|
||||
|
|
@ -253,13 +253,13 @@
|
|||
"dist/human.esm-nobundle.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 199199
|
||||
"bytes": 194427
|
||||
},
|
||||
"dist/human.esm-nobundle.js": {
|
||||
"imports": [],
|
||||
"inputs": {
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytesInOutput": 3255
|
||||
"bytesInOutput": 3131
|
||||
},
|
||||
"src/facemesh/keypoints.js": {
|
||||
"bytesInOutput": 1950
|
||||
|
|
@ -280,13 +280,13 @@
|
|||
"bytesInOutput": 9995
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytesInOutput": 1391
|
||||
"bytesInOutput": 1259
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytesInOutput": 1142
|
||||
"bytesInOutput": 934
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 1147
|
||||
"bytesInOutput": 944
|
||||
},
|
||||
"src/posenet/modelBase.js": {
|
||||
"bytesInOutput": 597
|
||||
|
|
@ -313,10 +313,10 @@
|
|||
"bytesInOutput": 608
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytesInOutput": 1850
|
||||
"bytesInOutput": 1058
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytesInOutput": 885
|
||||
"bytesInOutput": 846
|
||||
},
|
||||
"src/posenet/posenet.js": {
|
||||
"bytesInOutput": 464
|
||||
|
|
@ -325,31 +325,31 @@
|
|||
"bytesInOutput": 1400
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytesInOutput": 2040
|
||||
"bytesInOutput": 2067
|
||||
},
|
||||
"src/handpose/keypoints.js": {
|
||||
"bytesInOutput": 160
|
||||
},
|
||||
"src/handpose/util.js": {
|
||||
"bytesInOutput": 984
|
||||
"bytesInOutput": 977
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytesInOutput": 3218
|
||||
"bytesInOutput": 3209
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytesInOutput": 1317
|
||||
"bytesInOutput": 1211
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 1146
|
||||
"bytesInOutput": 1130
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 2305
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 4135
|
||||
"bytesInOutput": 4349
|
||||
}
|
||||
},
|
||||
"bytes": 69965
|
||||
"bytes": 68570
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 4862,
|
||||
"bytes": 4762,
|
||||
"imports": []
|
||||
},
|
||||
"node_modules/@tensorflow/tfjs-backend-cpu/dist/tf-backend-cpu.node.js": {
|
||||
|
|
@ -153,7 +153,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2019,
|
||||
"bytes": 1646,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -161,7 +161,7 @@
|
|||
]
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytes": 7407,
|
||||
"bytes": 7161,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -177,7 +177,7 @@
|
|||
]
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytes": 2816,
|
||||
"bytes": 2568,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -241,7 +241,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytes": 4296,
|
||||
"bytes": 4277,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -252,7 +252,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytes": 2356,
|
||||
"bytes": 2152,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -273,7 +273,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytes": 8178,
|
||||
"bytes": 8169,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -291,7 +291,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 8299,
|
||||
"bytes": 8784,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -380,7 +380,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytes": 3472,
|
||||
"bytes": 3447,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -417,11 +417,8 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytes": 4202,
|
||||
"bytes": 2260,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
},
|
||||
{
|
||||
"path": "src/posenet/keypoints.js"
|
||||
}
|
||||
|
|
@ -436,7 +433,7 @@
|
|||
]
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytes": 1937,
|
||||
"bytes": 1574,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -464,7 +461,7 @@
|
|||
"dist/human.esm.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 4960250
|
||||
"bytes": 4955479
|
||||
},
|
||||
"dist/human.esm.js": {
|
||||
"imports": [],
|
||||
|
|
@ -527,7 +524,7 @@
|
|||
"bytesInOutput": 765
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytesInOutput": 3268
|
||||
"bytesInOutput": 3142
|
||||
},
|
||||
"src/facemesh/keypoints.js": {
|
||||
"bytesInOutput": 1951
|
||||
|
|
@ -548,13 +545,13 @@
|
|||
"bytesInOutput": 9996
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytesInOutput": 1376
|
||||
"bytesInOutput": 1242
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytesInOutput": 1143
|
||||
"bytesInOutput": 930
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 1142
|
||||
"bytesInOutput": 932
|
||||
},
|
||||
"src/posenet/modelBase.js": {
|
||||
"bytesInOutput": 575
|
||||
|
|
@ -581,10 +578,10 @@
|
|||
"bytesInOutput": 609
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytesInOutput": 1840
|
||||
"bytesInOutput": 1067
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytesInOutput": 863
|
||||
"bytesInOutput": 824
|
||||
},
|
||||
"src/posenet/posenet.js": {
|
||||
"bytesInOutput": 479
|
||||
|
|
@ -593,7 +590,7 @@
|
|||
"bytesInOutput": 1386
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytesInOutput": 2050
|
||||
"bytesInOutput": 2073
|
||||
},
|
||||
"src/handpose/keypoints.js": {
|
||||
"bytesInOutput": 161
|
||||
|
|
@ -602,22 +599,22 @@
|
|||
"bytesInOutput": 993
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytesInOutput": 3214
|
||||
"bytesInOutput": 3205
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytesInOutput": 1303
|
||||
"bytesInOutput": 1189
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 1147
|
||||
"bytesInOutput": 1131
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 2306
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 4246
|
||||
"bytesInOutput": 4464
|
||||
}
|
||||
},
|
||||
"bytes": 1106891
|
||||
"bytes": 1105498
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 4862,
|
||||
"bytes": 4762,
|
||||
"imports": []
|
||||
},
|
||||
"node_modules/@tensorflow/tfjs-backend-cpu/dist/tf-backend-cpu.node.js": {
|
||||
|
|
@ -153,7 +153,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2019,
|
||||
"bytes": 1646,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -161,7 +161,7 @@
|
|||
]
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytes": 7407,
|
||||
"bytes": 7161,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -177,7 +177,7 @@
|
|||
]
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytes": 2816,
|
||||
"bytes": 2568,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -241,7 +241,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytes": 4296,
|
||||
"bytes": 4277,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -252,7 +252,7 @@
|
|||
]
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytes": 2356,
|
||||
"bytes": 2152,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -273,7 +273,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytes": 8178,
|
||||
"bytes": 8169,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -291,7 +291,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 8299,
|
||||
"bytes": 8784,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -380,7 +380,7 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytes": 3472,
|
||||
"bytes": 3447,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -417,11 +417,8 @@
|
|||
]
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytes": 4202,
|
||||
"bytes": 2260,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
},
|
||||
{
|
||||
"path": "src/posenet/keypoints.js"
|
||||
}
|
||||
|
|
@ -436,7 +433,7 @@
|
|||
]
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytes": 1937,
|
||||
"bytes": 1574,
|
||||
"imports": [
|
||||
{
|
||||
"path": "node_modules/@tensorflow/tfjs/dist/tf.node.js"
|
||||
|
|
@ -464,7 +461,7 @@
|
|||
"dist/human.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 4960250
|
||||
"bytes": 4955479
|
||||
},
|
||||
"dist/human.js": {
|
||||
"imports": [],
|
||||
|
|
@ -527,7 +524,7 @@
|
|||
"bytesInOutput": 765
|
||||
},
|
||||
"src/facemesh/blazeface.js": {
|
||||
"bytesInOutput": 3268
|
||||
"bytesInOutput": 3142
|
||||
},
|
||||
"src/facemesh/keypoints.js": {
|
||||
"bytesInOutput": 1951
|
||||
|
|
@ -548,13 +545,13 @@
|
|||
"bytesInOutput": 9996
|
||||
},
|
||||
"src/facemesh/facemesh.js": {
|
||||
"bytesInOutput": 1376
|
||||
"bytesInOutput": 1242
|
||||
},
|
||||
"src/ssrnet/ssrnet.js": {
|
||||
"bytesInOutput": 1143
|
||||
"bytesInOutput": 930
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 1142
|
||||
"bytesInOutput": 932
|
||||
},
|
||||
"src/posenet/modelBase.js": {
|
||||
"bytesInOutput": 575
|
||||
|
|
@ -581,10 +578,10 @@
|
|||
"bytesInOutput": 609
|
||||
},
|
||||
"src/posenet/util.js": {
|
||||
"bytesInOutput": 1840
|
||||
"bytesInOutput": 1067
|
||||
},
|
||||
"src/posenet/modelPoseNet.js": {
|
||||
"bytesInOutput": 863
|
||||
"bytesInOutput": 824
|
||||
},
|
||||
"src/posenet/posenet.js": {
|
||||
"bytesInOutput": 479
|
||||
|
|
@ -593,7 +590,7 @@
|
|||
"bytesInOutput": 1386
|
||||
},
|
||||
"src/handpose/handdetector.js": {
|
||||
"bytesInOutput": 2050
|
||||
"bytesInOutput": 2073
|
||||
},
|
||||
"src/handpose/keypoints.js": {
|
||||
"bytesInOutput": 161
|
||||
|
|
@ -602,22 +599,22 @@
|
|||
"bytesInOutput": 993
|
||||
},
|
||||
"src/handpose/pipeline.js": {
|
||||
"bytesInOutput": 3214
|
||||
"bytesInOutput": 3205
|
||||
},
|
||||
"src/handpose/handpose.js": {
|
||||
"bytesInOutput": 1303
|
||||
"bytesInOutput": 1189
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 1147
|
||||
"bytesInOutput": 1131
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 2306
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 4246
|
||||
"bytesInOutput": 4464
|
||||
}
|
||||
},
|
||||
"bytes": 1106900
|
||||
"bytes": 1105507
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -6,16 +6,6 @@ let last = [];
|
|||
let frame = 0;
|
||||
const multiplier = 1.5;
|
||||
|
||||
function getImage(image, size) {
|
||||
const tensor = tf.tidy(() => {
|
||||
const buffer = tf.browser.fromPixels(image, 1);
|
||||
const resize = tf.image.resizeBilinear(buffer, [size, size]);
|
||||
const expand = tf.cast(tf.expandDims(resize, 0), 'float32');
|
||||
return expand;
|
||||
});
|
||||
return tensor;
|
||||
}
|
||||
|
||||
async function load(config) {
|
||||
if (!models.emotion) models.emotion = await tf.loadGraphModel(config.face.emotion.modelPath);
|
||||
return models.emotion;
|
||||
|
|
@ -27,25 +17,23 @@ async function predict(image, config) {
|
|||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
const enhance = tf.tidy(() => {
|
||||
if (image instanceof tf.Tensor) {
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.emotion.inputSize, config.face.emotion.inputSize], false);
|
||||
const [r, g, b] = tf.split(resize, 3, 3);
|
||||
if (config.face.emotion.useGrayscale) {
|
||||
// weighted rgb to grayscale: https://www.mathworks.com/help/matlab/ref/rgb2gray.html
|
||||
const r1 = tf.mul(r, [0.2989]);
|
||||
const g1 = tf.mul(g, [0.5870]);
|
||||
const b1 = tf.mul(b, [0.1140]);
|
||||
const grayscale = tf.addN([r1, g1, b1]);
|
||||
return grayscale;
|
||||
}
|
||||
return g;
|
||||
}
|
||||
return getImage(image, config.face.emotion.inputSize);
|
||||
});
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.emotion.inputSize, config.face.emotion.inputSize], false);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
// weighted rgb to grayscale: https://www.mathworks.com/help/matlab/ref/rgb2gray.html
|
||||
const redNorm = tf.mul(red, [0.2989]);
|
||||
const greenNorm = tf.mul(green, [0.5870]);
|
||||
const blueNorm = tf.mul(blue, [0.1140]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const obj = [];
|
||||
if (config.face.emotion.enabled) {
|
||||
const emotionT = await models.emotion.predict(enhance);
|
||||
const emotionT = await models.emotion.predict(grayscale);
|
||||
const data = await emotionT.data();
|
||||
for (let i = 0; i < data.length; i++) {
|
||||
if (multiplier * data[i] > config.face.emotion.minConfidence) obj.push({ score: Math.min(0.99, Math.trunc(100 * multiplier * data[i]) / 100), emotion: annotations[i] });
|
||||
|
|
@ -53,7 +41,7 @@ async function predict(image, config) {
|
|||
obj.sort((a, b) => b.score - a.score);
|
||||
tf.dispose(emotionT);
|
||||
}
|
||||
tf.dispose(enhance);
|
||||
tf.dispose(grayscale);
|
||||
last = obj;
|
||||
return obj;
|
||||
}
|
||||
|
|
|
|||
|
|
@ -144,13 +144,7 @@ class BlazeFaceModel {
|
|||
}
|
||||
|
||||
async estimateFaces(input) {
|
||||
const imageRaw = !(input instanceof tf.Tensor) ? tf.browser.fromPixels(input) : input;
|
||||
const imageCast = imageRaw.toFloat();
|
||||
const image = imageCast.expandDims(0);
|
||||
imageRaw.dispose();
|
||||
imageCast.dispose();
|
||||
const { boxes, scaleFactor } = await this.getBoundingBoxes(image);
|
||||
image.dispose();
|
||||
const { boxes, scaleFactor } = await this.getBoundingBoxes(input);
|
||||
return Promise.all(boxes.map(async (face) => {
|
||||
const scaledBox = scaleBoxFromPrediction(face, scaleFactor);
|
||||
const [landmarkData, boxData, probabilityData] = await Promise.all([face.landmarks, scaledBox, face.probability].map(async (d) => d.array()));
|
||||
|
|
|
|||
|
|
@ -13,13 +13,7 @@ class MediaPipeFaceMesh {
|
|||
|
||||
async estimateFaces(input, config) {
|
||||
if (config) this.config = config;
|
||||
const imageRaw = !(input instanceof tf.Tensor) ? tf.browser.fromPixels(input) : input;
|
||||
const imageCast = imageRaw.toFloat();
|
||||
const image = imageCast.expandDims(0);
|
||||
imageRaw.dispose();
|
||||
imageCast.dispose();
|
||||
const predictions = await this.pipeline.predict(image, config);
|
||||
tf.dispose(image);
|
||||
const predictions = await this.pipeline.predict(input, config);
|
||||
const results = [];
|
||||
for (const prediction of (predictions || [])) {
|
||||
// guard against disposed tensors on long running operations such as pause in middle of processing
|
||||
|
|
|
|||
|
|
@ -32,9 +32,9 @@ class HandDetector {
|
|||
}
|
||||
|
||||
async getBoundingBoxes(input) {
|
||||
const normalizedInput = tf.tidy(() => tf.mul(tf.sub(input, 0.5), 2));
|
||||
const batchedPrediction = this.model.predict(normalizedInput);
|
||||
const batchedPrediction = this.model.predict(input);
|
||||
const prediction = batchedPrediction.squeeze();
|
||||
console.log(prediction);
|
||||
// Regression score for each anchor point.
|
||||
const scores = tf.tidy(() => tf.sigmoid(tf.slice(prediction, [0, 0], [-1, 1])).squeeze());
|
||||
// Bounding box for each anchor point.
|
||||
|
|
@ -42,11 +42,7 @@ class HandDetector {
|
|||
const boxes = this.normalizeBoxes(rawBoxes);
|
||||
const boxesWithHandsTensor = await tf.image.nonMaxSuppressionAsync(boxes, scores, this.maxHands, this.iouThreshold, this.scoreThreshold);
|
||||
const boxesWithHands = await boxesWithHandsTensor.array();
|
||||
const toDispose = [normalizedInput, batchedPrediction, boxesWithHandsTensor, prediction, boxes, rawBoxes, scores];
|
||||
// if (boxesWithHands.length === 0) {
|
||||
// toDispose.forEach((tensor) => tensor.dispose());
|
||||
// return null;
|
||||
// }
|
||||
const toDispose = [batchedPrediction, boxesWithHandsTensor, prediction, boxes, rawBoxes, scores];
|
||||
const detectedHands = tf.tidy(() => {
|
||||
const detectedBoxes = [];
|
||||
for (const i in boxesWithHands) {
|
||||
|
|
@ -69,12 +65,18 @@ class HandDetector {
|
|||
* @param input The image to classify.
|
||||
*/
|
||||
async estimateHandBounds(input, config) {
|
||||
const inputHeight = input.shape[1];
|
||||
const inputWidth = input.shape[2];
|
||||
// const inputHeight = input.shape[2];
|
||||
// const inputWidth = input.shape[1];
|
||||
this.iouThreshold = config.iouThreshold;
|
||||
this.scoreThreshold = config.scoreThreshold;
|
||||
this.maxHands = config.maxHands;
|
||||
const image = tf.tidy(() => input.resizeBilinear([this.width, this.height]).div(255));
|
||||
const resized = input.resizeBilinear([this.width, this.height]);
|
||||
const divided = resized.div(255);
|
||||
const normalized = divided.sub(0.5);
|
||||
const image = normalized.mul(2.0);
|
||||
resized.dispose();
|
||||
divided.dispose();
|
||||
normalized.dispose();
|
||||
const predictions = await this.getBoundingBoxes(image);
|
||||
image.dispose();
|
||||
if (!predictions || (predictions.length === 0)) return null;
|
||||
|
|
@ -87,7 +89,7 @@ class HandDetector {
|
|||
const palmLandmarks = await prediction.palmLandmarks.array();
|
||||
prediction.boxes.dispose();
|
||||
prediction.palmLandmarks.dispose();
|
||||
hands.push(bounding.scaleBoxCoordinates({ startPoint, endPoint, palmLandmarks }, [inputWidth / this.width, inputHeight / this.height]));
|
||||
hands.push(bounding.scaleBoxCoordinates({ startPoint, endPoint, palmLandmarks }, [input.shape[2] / this.width, input.shape[1] / this.height]));
|
||||
}
|
||||
return hands;
|
||||
}
|
||||
|
|
|
|||
|
|
@ -12,14 +12,7 @@ class HandPose {
|
|||
this.skipFrames = config.skipFrames;
|
||||
this.detectionConfidence = config.minConfidence;
|
||||
this.maxHands = config.maxHands;
|
||||
const image = tf.tidy(() => {
|
||||
if (!(input instanceof tf.Tensor)) {
|
||||
input = tf.browser.fromPixels(input);
|
||||
}
|
||||
return input.toFloat().expandDims(0);
|
||||
});
|
||||
const predictions = await this.pipeline.estimateHands(image, config);
|
||||
image.dispose();
|
||||
const predictions = await this.pipeline.estimateHands(input, config);
|
||||
const hands = [];
|
||||
if (!predictions) return hands;
|
||||
for (const prediction of predictions) {
|
||||
|
|
|
|||
|
|
@ -76,7 +76,7 @@ class HandPipeline {
|
|||
}
|
||||
|
||||
async estimateHands(image, config) {
|
||||
this.maxContinuousChecks = config.skipFrames;
|
||||
this.skipFrames = config.skipFrames;
|
||||
this.detectionConfidence = config.minConfidence;
|
||||
this.maxHands = config.maxHands;
|
||||
this.runsWithoutHandDetector++;
|
||||
|
|
|
|||
29
src/human.js
29
src/human.js
|
|
@ -71,7 +71,9 @@ function mergeDeep(...objects) {
|
|||
|
||||
function sanity(input) {
|
||||
if (!input) return 'input is not defined';
|
||||
if (tf.ENV.flags.IS_BROWSER && (input instanceof ImageData || input instanceof HTMLImageElement || input instanceof HTMLCanvasElement || input instanceof HTMLVideoElement || input instanceof HTMLMediaElement)) {
|
||||
if (!(input instanceof tf.Tensor)
|
||||
|| (tf.ENV.flags.IS_BROWSER
|
||||
&& (input instanceof ImageData || input instanceof HTMLImageElement || input instanceof HTMLCanvasElement || input instanceof HTMLVideoElement || input instanceof HTMLMediaElement))) {
|
||||
const width = input.naturalWidth || input.videoWidth || input.width || (input.shape && (input.shape[1] > 0));
|
||||
if (!width || (width === 0)) return 'input is empty';
|
||||
}
|
||||
|
|
@ -99,6 +101,20 @@ async function load(userConfig) {
|
|||
if (config.face.enabled && config.face.emotion.enabled && !models.emotion) models.emotion = await emotion.load(config);
|
||||
}
|
||||
|
||||
function tfImage(input) {
|
||||
let image;
|
||||
if (input instanceof tf.Tensor) {
|
||||
image = tf.clone(input);
|
||||
} else {
|
||||
const pixels = tf.browser.fromPixels(input);
|
||||
const casted = pixels.toFloat();
|
||||
image = casted.expandDims(0);
|
||||
pixels.dispose();
|
||||
casted.dispose();
|
||||
}
|
||||
return image;
|
||||
}
|
||||
|
||||
async function detect(input, userConfig = {}) {
|
||||
state = 'config';
|
||||
const perf = {};
|
||||
|
|
@ -151,11 +167,13 @@ async function detect(input, userConfig = {}) {
|
|||
|
||||
analyze('Start Detect:');
|
||||
|
||||
const imageTensor = tfImage(input);
|
||||
|
||||
// run posenet
|
||||
state = 'run:body';
|
||||
timeStamp = now();
|
||||
analyze('Start PoseNet');
|
||||
const poseRes = config.body.enabled ? await models.posenet.estimatePoses(input, config.body) : [];
|
||||
const poseRes = config.body.enabled ? await models.posenet.estimatePoses(imageTensor, config.body) : [];
|
||||
analyze('End PoseNet:');
|
||||
perf.body = Math.trunc(now() - timeStamp);
|
||||
|
||||
|
|
@ -163,7 +181,7 @@ async function detect(input, userConfig = {}) {
|
|||
state = 'run:hand';
|
||||
timeStamp = now();
|
||||
analyze('Start HandPose:');
|
||||
const handRes = config.hand.enabled ? await models.handpose.estimateHands(input, config.hand) : [];
|
||||
const handRes = config.hand.enabled ? await models.handpose.estimateHands(imageTensor, config.hand) : [];
|
||||
analyze('End HandPose:');
|
||||
perf.hand = Math.trunc(now() - timeStamp);
|
||||
|
||||
|
|
@ -173,7 +191,7 @@ async function detect(input, userConfig = {}) {
|
|||
state = 'run:face';
|
||||
timeStamp = now();
|
||||
analyze('Start FaceMesh:');
|
||||
const faces = await models.facemesh.estimateFaces(input, config.face);
|
||||
const faces = await models.facemesh.estimateFaces(imageTensor, config.face);
|
||||
perf.face = Math.trunc(now() - timeStamp);
|
||||
for (const face of faces) {
|
||||
// is something went wrong, skip the face
|
||||
|
|
@ -210,10 +228,11 @@ async function detect(input, userConfig = {}) {
|
|||
emotion: emotionData,
|
||||
iris: (iris !== 0) ? Math.trunc(100 * 11.7 /* human iris size in mm */ / iris) / 100 : 0,
|
||||
});
|
||||
analyze('End FaceMesh:');
|
||||
}
|
||||
analyze('End FaceMesh:');
|
||||
}
|
||||
|
||||
imageTensor.dispose();
|
||||
state = 'idle';
|
||||
|
||||
if (config.scoped) tf.engine().endScope();
|
||||
|
|
|
|||
|
|
@ -31,8 +31,9 @@ class PoseNet {
|
|||
async estimatePoses(input, config) {
|
||||
const outputStride = config.outputStride;
|
||||
// const inputResolution = config.inputResolution;
|
||||
const [height, width] = util.getInputTensorDimensions(input);
|
||||
const { resized, padding } = util.padAndResizeTo(input, [config.inputResolution, config.inputResolution]);
|
||||
const height = input.shape[1];
|
||||
const width = input.shape[2];
|
||||
const resized = util.resizeTo(input, [config.inputResolution, config.inputResolution]);
|
||||
const { heatmapScores, offsets, displacementFwd, displacementBwd } = this.baseModel.predict(resized);
|
||||
const allTensorBuffers = await util.toTensorBuffers3D([heatmapScores, offsets, displacementFwd, displacementBwd]);
|
||||
const scoresBuffer = allTensorBuffers[0];
|
||||
|
|
@ -40,7 +41,7 @@ class PoseNet {
|
|||
const displacementsFwdBuffer = allTensorBuffers[2];
|
||||
const displacementsBwdBuffer = allTensorBuffers[3];
|
||||
const poses = await decodeMultiple.decodeMultiplePoses(scoresBuffer, offsetsBuffer, displacementsFwdBuffer, displacementsBwdBuffer, outputStride, config.maxDetections, config.scoreThreshold, config.nmsRadius);
|
||||
const resultPoses = util.scaleAndFlipPoses(poses, [height, width], [config.inputResolution, config.inputResolution], padding);
|
||||
const resultPoses = util.scaleAndFlipPoses(poses, [height, width], [config.inputResolution, config.inputResolution]);
|
||||
heatmapScores.dispose();
|
||||
offsets.dispose();
|
||||
displacementFwd.dispose();
|
||||
|
|
|
|||
|
|
@ -1,4 +1,3 @@
|
|||
const tf = require('@tensorflow/tfjs');
|
||||
const kpt = require('./keypoints');
|
||||
|
||||
function eitherPointDoesntMeetConfidence(a, b, minConfidence) {
|
||||
|
|
@ -41,78 +40,28 @@ async function toTensorBuffers3D(tensors) {
|
|||
}
|
||||
exports.toTensorBuffers3D = toTensorBuffers3D;
|
||||
|
||||
function scalePose(pose, scaleY, scaleX, offsetY = 0, offsetX = 0) {
|
||||
function scalePose(pose, scaleY, scaleX) {
|
||||
return {
|
||||
score: pose.score,
|
||||
keypoints: pose.keypoints.map(({ score, part, position }) => ({
|
||||
score,
|
||||
part,
|
||||
position: {
|
||||
x: position.x * scaleX + offsetX,
|
||||
y: position.y * scaleY + offsetY,
|
||||
},
|
||||
position: { x: position.x * scaleX, y: position.y * scaleY },
|
||||
})),
|
||||
};
|
||||
}
|
||||
exports.scalePose = scalePose;
|
||||
|
||||
function scalePoses(poses, scaleY, scaleX, offsetY = 0, offsetX = 0) {
|
||||
if (scaleX === 1 && scaleY === 1 && offsetY === 0 && offsetX === 0) {
|
||||
return poses;
|
||||
}
|
||||
return poses.map((pose) => scalePose(pose, scaleY, scaleX, offsetY, offsetX));
|
||||
function resizeTo(image, [targetH, targetW]) {
|
||||
const input = image.squeeze(0);
|
||||
const resized = input.resizeBilinear([targetH, targetW]);
|
||||
input.dispose();
|
||||
return resized;
|
||||
}
|
||||
exports.scalePoses = scalePoses;
|
||||
exports.resizeTo = resizeTo;
|
||||
|
||||
function getInputTensorDimensions(input) {
|
||||
return input instanceof tf.Tensor ? [input.shape[0], input.shape[1]] : [input.height, input.width];
|
||||
}
|
||||
exports.getInputTensorDimensions = getInputTensorDimensions;
|
||||
|
||||
function toInputTensor(input) {
|
||||
return input instanceof tf.Tensor ? input : tf.browser.fromPixels(input);
|
||||
}
|
||||
exports.toInputTensor = toInputTensor;
|
||||
|
||||
function toResizedInputTensor(input, resizeHeight, resizeWidth) {
|
||||
return tf.tidy(() => {
|
||||
const imageTensor = toInputTensor(input);
|
||||
return imageTensor.resizeBilinear([resizeHeight, resizeWidth]);
|
||||
});
|
||||
}
|
||||
exports.toResizedInputTensor = toResizedInputTensor;
|
||||
|
||||
function padAndResizeTo(input, [targetH, targetW]) {
|
||||
const [height, width] = getInputTensorDimensions(input);
|
||||
const targetAspect = targetW / targetH;
|
||||
const aspect = width / height;
|
||||
let [padT, padB, padL, padR] = [0, 0, 0, 0];
|
||||
if (aspect < targetAspect) {
|
||||
// pads the width
|
||||
padT = 0;
|
||||
padB = 0;
|
||||
padL = Math.round(0.5 * (targetAspect * height - width));
|
||||
padR = Math.round(0.5 * (targetAspect * height - width));
|
||||
} else {
|
||||
// pads the height
|
||||
padT = Math.round(0.5 * ((1.0 / targetAspect) * width - height));
|
||||
padB = Math.round(0.5 * ((1.0 / targetAspect) * width - height));
|
||||
padL = 0;
|
||||
padR = 0;
|
||||
}
|
||||
const resized = tf.tidy(() => {
|
||||
let imageTensor = toInputTensor(input);
|
||||
imageTensor = tf.pad3d(imageTensor, [[padT, padB], [padL, padR], [0, 0]]);
|
||||
return imageTensor.resizeBilinear([targetH, targetW]);
|
||||
});
|
||||
return { resized, padding: { top: padT, left: padL, right: padR, bottom: padB } };
|
||||
}
|
||||
exports.padAndResizeTo = padAndResizeTo;
|
||||
|
||||
function scaleAndFlipPoses(poses, [height, width], [inputResolutionHeight, inputResolutionWidth], padding) {
|
||||
const scaleY = (height + padding.top + padding.bottom) / (inputResolutionHeight);
|
||||
const scaleX = (width + padding.left + padding.right) / (inputResolutionWidth);
|
||||
const scaledPoses = scalePoses(poses, scaleY, scaleX, -padding.top, -padding.left);
|
||||
function scaleAndFlipPoses(poses, [height, width], [inputResolutionHeight, inputResolutionWidth]) {
|
||||
const scaledPoses = poses.map((pose) => scalePose(pose, height / inputResolutionHeight, width / inputResolutionWidth));
|
||||
return scaledPoses;
|
||||
}
|
||||
exports.scaleAndFlipPoses = scaleAndFlipPoses;
|
||||
|
|
|
|||
|
|
@ -4,13 +4,6 @@ const models = {};
|
|||
let last = { age: 0, gender: '' };
|
||||
let frame = 0;
|
||||
|
||||
async function getImage(image, size) {
|
||||
const buffer = tf.browser.fromPixels(image);
|
||||
const resize = tf.image.resizeBilinear(buffer, [size, size]);
|
||||
const expand = tf.cast(tf.expandDims(resize, 0), 'float32');
|
||||
return expand;
|
||||
}
|
||||
|
||||
async function loadAge(config) {
|
||||
if (!models.age) models.age = await tf.loadGraphModel(config.face.age.modelPath);
|
||||
return models.age;
|
||||
|
|
@ -27,14 +20,9 @@ async function predict(image, config) {
|
|||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
let enhance;
|
||||
if (image instanceof tf.Tensor) {
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.age.inputSize, config.face.age.inputSize], false);
|
||||
enhance = tf.mul(resize, [255.0]);
|
||||
tf.dispose(resize);
|
||||
} else {
|
||||
enhance = await getImage(image, config.face.age.inputSize);
|
||||
}
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.age.inputSize, config.face.age.inputSize], false);
|
||||
const enhance = tf.mul(resize, [255.0]);
|
||||
tf.dispose(resize);
|
||||
|
||||
const promises = [];
|
||||
let ageT;
|
||||
|
|
|
|||
Loading…
Reference in New Issue