mirror of https://github.com/vladmandic/human
full rebuild
Vladimir Mandic
parent
8d38c977be
commit
5c3de60f44
18
README.md
18
README.md
|
|
@ -1,6 +1,6 @@
|
|||
# Human Library
|
||||
|
||||
## 3D Face Detection, Body Pose, Hand & Finger Tracking, Iris Tracking, Age & Gender Prediction, Emotion Prediction & Gesture Recognition
|
||||
## 3D Face Detection, Face Embedding & Recognition, Body Pose Tracking, Hand & Finger Tracking, Iris Analysis, Age & Gender & Emotion Prediction & Gesture Recognition
|
||||
|
||||
<br>
|
||||
|
||||
|
|
@ -12,18 +12,20 @@
|
|||
- [**Issues Tracker**](https://github.com/vladmandic/human/issues)
|
||||
- [**Change Log**](https://github.com/vladmandic/human/wiki/Change-Log)
|
||||
|
||||
<br>
|
||||
|
||||
### Wiki pages:
|
||||
|
||||
- [**Home**](https://github.com/vladmandic/human/wiki)
|
||||
- [**Demos**](https://github.com/vladmandic/human/wiki/Demos)
|
||||
- [**Installation**](https://github.com/vladmandic/human/wiki/Install)
|
||||
- [**Usage**](https://github.com/vladmandic/human/wiki/Usage)
|
||||
- [**Configuration**](https://github.com/vladmandic/human/wiki/Configuration)
|
||||
- [**Outputs**](https://github.com/vladmandic/human/wiki/Outputs)
|
||||
- [**Notes on Backends**](https://github.com/vladmandic/human/wiki/Backends)
|
||||
- [**Usage & Functions**](https://github.com/vladmandic/human/wiki/Usage)
|
||||
- [**Configuration Details**](https://github.com/vladmandic/human/wiki/Configuration)
|
||||
- [**Output Details**](https://github.com/vladmandic/human/wiki/Outputs)
|
||||
- [**Face Embedding and Recognition**](https://github.com/vladmandic/human/wiki/Embedding)
|
||||
- [**Gesture Recognition**](https://github.com/vladmandic/human/wiki/Gesture)
|
||||
|
||||
### Additional notes:
|
||||
|
||||
- [**Notes on Backends**](https://github.com/vladmandic/human/wiki/Backends)
|
||||
- [**Development Server**](https://github.com/vladmandic/human/wiki/Development-Server)
|
||||
- [**Build Process**](https://github.com/vladmandic/human/wiki/Build-Process)
|
||||
- [**List of Models**](https://github.com/vladmandic/human/wiki/Models)
|
||||
|
|
@ -34,7 +36,7 @@
|
|||
<br>
|
||||
|
||||
Compatible with *Browser*, *WebWorker* and *NodeJS* execution on both Windows and Linux
|
||||
- Browser/WebWorker: Compatible with *CPU*, *WebGL* and *WASM* backends
|
||||
- Browser/WebWorker: Compatible with *CPU*, *WebGL*, *WASM* and *WebGPU* backends
|
||||
- NodeJS: Compatible with software *tfjs-node* and CUDA accelerated backends *tfjs-node-gpu*
|
||||
- (and maybe with React-Native as it doesn't use any DOM objects)
|
||||
|
||||
|
|
|
|||
|
|
@ -4270,15 +4270,15 @@ var require_facepipeline = __commonJS((exports) => {
|
|||
rotatedImage = tf.image.rotateWithOffset(input, angle, 0, faceCenterNormalized);
|
||||
rotationMatrix = util.buildRotationMatrix(-angle, faceCenter);
|
||||
}
|
||||
const boxCPU = {startPoint: box.startPoint, endPoint: box.endPoint};
|
||||
const face2 = bounding.cutBoxFromImageAndResize(boxCPU, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const face2 = bounding.cutBoxFromImageAndResize({startPoint: box.startPoint, endPoint: box.endPoint}, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const outputFace = config2.detector.rotation ? tf.image.rotateWithOffset(face2, angle) : face2;
|
||||
if (!config2.mesh.enabled) {
|
||||
const prediction2 = {
|
||||
coords: null,
|
||||
box,
|
||||
faceConfidence: null,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
return prediction2;
|
||||
}
|
||||
|
|
@ -4323,7 +4323,7 @@ var require_facepipeline = __commonJS((exports) => {
|
|||
box: landmarksBox,
|
||||
faceConfidence: confidenceVal,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
this.storedBoxes[i] = {...landmarksBox, landmarks: transformedCoords.arraySync(), confidence: box.confidence, faceConfidence: confidenceVal};
|
||||
return prediction;
|
||||
|
|
@ -4435,7 +4435,6 @@ var require_age = __commonJS((exports) => {
|
|||
const models = {};
|
||||
let last = {age: 0};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
async function load2(config2) {
|
||||
if (!models.age) {
|
||||
models.age = await loadGraphModel(config2.face.age.modelPath);
|
||||
|
|
@ -4444,19 +4443,15 @@ var require_age = __commonJS((exports) => {
|
|||
return models.age;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.age)
|
||||
return null;
|
||||
if (frame < config2.face.age.skipFrames && last.age && last.age > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.age.inputSize, config2.face.age.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.age.inputSize, config2.face.age.inputSize], false);
|
||||
const enhance = tf.mul(resize, [255]);
|
||||
tf.dispose(resize);
|
||||
let ageT;
|
||||
|
|
@ -4491,7 +4486,6 @@ var require_gender = __commonJS((exports) => {
|
|||
let last = {gender: ""};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
let alternative = false;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.gender) {
|
||||
|
|
@ -4502,19 +4496,15 @@ var require_gender = __commonJS((exports) => {
|
|||
return models.gender;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.gender)
|
||||
return null;
|
||||
if (frame < config2.face.gender.skipFrames && last.gender !== "") {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.gender.inputSize, config2.face.gender.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.gender.inputSize, config2.face.gender.inputSize], false);
|
||||
let enhance;
|
||||
if (alternative) {
|
||||
enhance = tf.tidy(() => {
|
||||
|
|
@ -4573,7 +4563,6 @@ var require_emotion = __commonJS((exports) => {
|
|||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
const scale = 1;
|
||||
async function load2(config2) {
|
||||
|
|
@ -4584,19 +4573,15 @@ var require_emotion = __commonJS((exports) => {
|
|||
return models.emotion;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.emotion)
|
||||
return null;
|
||||
if (frame < config2.face.emotion.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.emotion.inputSize, config2.face.emotion.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.emotion.inputSize, config2.face.emotion.inputSize], false);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
|
|
@ -4643,10 +4628,6 @@ var require_emotion = __commonJS((exports) => {
|
|||
var require_embedding = __commonJS((exports) => {
|
||||
const profile2 = __toModule(require_profile());
|
||||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.embedding) {
|
||||
models.embedding = await loadGraphModel(config2.face.embedding.modelPath);
|
||||
|
|
@ -4654,57 +4635,37 @@ var require_embedding = __commonJS((exports) => {
|
|||
}
|
||||
return models.embedding;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (frame < config2.face.embedding.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
function simmilarity2(embedding1, embedding2) {
|
||||
if ((embedding1 == null ? void 0 : embedding1.length) !== (embedding2 == null ? void 0 : embedding2.length))
|
||||
return 0;
|
||||
const distance = 10 * Math.sqrt(embedding1.map((val, i) => val - embedding2[i]).reduce((dist2, diff) => dist2 + diff ** 2, 0));
|
||||
const confidence = 2 * (0.5 - distance);
|
||||
return Math.trunc(1e3 * confidence) / 1e3;
|
||||
}
|
||||
frame = 0;
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.embedding)
|
||||
return null;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.embedding.inputSize, config2.face.embedding.inputSize]);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
const greenNorm = tf.mul(green, rgb[1]);
|
||||
const blueNorm = tf.mul(blue, rgb[2]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const normalize = tf.tidy(() => grayscale.sub(0.5).mul(2));
|
||||
grayscale.dispose();
|
||||
const obj = [];
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.embedding.inputSize, config2.face.embedding.inputSize], false);
|
||||
let data2 = [];
|
||||
if (config2.face.embedding.enabled) {
|
||||
let data2;
|
||||
if (!config2.profile) {
|
||||
console.log("model", models.embedding);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: normalize});
|
||||
data2 = embeddingT.dataSync();
|
||||
console.log("embedding", data2);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: resize});
|
||||
data2 = [...embeddingT.dataSync()];
|
||||
tf.dispose(embeddingT);
|
||||
} else {
|
||||
const profileData = await tf.profile(() => models.embedding.predict(normalize));
|
||||
data2 = profileData.result.dataSync();
|
||||
const profileData = await tf.profile(() => models.embedding.predict({img_inputs: resize}));
|
||||
data2 = [...profileData.result.dataSync()];
|
||||
profileData.result.dispose();
|
||||
profile2.run("emotion", profileData);
|
||||
}
|
||||
obj.sort((a, b) => b.score - a.score);
|
||||
}
|
||||
normalize.dispose();
|
||||
last = obj;
|
||||
resolve(obj);
|
||||
resize.dispose();
|
||||
resolve(data2);
|
||||
});
|
||||
}
|
||||
exports.predict = predict2;
|
||||
exports.simmilarity = simmilarity2;
|
||||
exports.load = load2;
|
||||
});
|
||||
|
||||
|
|
@ -24353,6 +24314,7 @@ var config_default = {
|
|||
detector: {
|
||||
modelPath: "../models/blazeface-back.json",
|
||||
inputSize: 256,
|
||||
rotation: false,
|
||||
maxFaces: 10,
|
||||
skipFrames: 15,
|
||||
minConfidence: 0.5,
|
||||
|
|
@ -24390,7 +24352,7 @@ var config_default = {
|
|||
modelPath: "../models/emotion-large.json"
|
||||
},
|
||||
embedding: {
|
||||
enabled: true,
|
||||
enabled: false,
|
||||
inputSize: 112,
|
||||
modelPath: "../models/mobilefacenet.json"
|
||||
}
|
||||
|
|
@ -24513,6 +24475,11 @@ class Human {
|
|||
}
|
||||
return null;
|
||||
}
|
||||
simmilarity(embedding1, embedding2) {
|
||||
if (this.config.face.embedding.enabled)
|
||||
return embedding.simmilarity(embedding1, embedding2);
|
||||
return 0;
|
||||
}
|
||||
async load(userConfig) {
|
||||
this.state = "load";
|
||||
const timeStamp = now();
|
||||
|
|
@ -24686,6 +24653,7 @@ class Human {
|
|||
return process3.canvas;
|
||||
}
|
||||
async detect(input, userConfig = {}) {
|
||||
return new Promise(async (resolve) => {
|
||||
this.state = "config";
|
||||
let timeStamp;
|
||||
this.config = mergeDeep(this.config, userConfig);
|
||||
|
|
@ -24695,9 +24663,8 @@ class Human {
|
|||
const error = this.sanity(input);
|
||||
if (error) {
|
||||
this.log(error, input);
|
||||
return {error};
|
||||
resolve({error});
|
||||
}
|
||||
return new Promise(async (resolve) => {
|
||||
let poseRes;
|
||||
let handRes;
|
||||
let faceRes;
|
||||
|
|
@ -24765,10 +24732,12 @@ class Human {
|
|||
resolve({face: faceRes, body: poseRes, hand: handRes, gesture: gestureRes, performance: this.perf, canvas: process3.canvas});
|
||||
});
|
||||
}
|
||||
async warmup(userConfig) {
|
||||
const warmup = new ImageData(255, 255);
|
||||
await this.detect(warmup, userConfig);
|
||||
async warmup(userConfig, sample) {
|
||||
if (!sample)
|
||||
sample = new ImageData(255, 255);
|
||||
const warmup = await this.detect(sample, userConfig);
|
||||
this.log("warmed up");
|
||||
return warmup;
|
||||
}
|
||||
}
|
||||
export {
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
|
|
@ -1,15 +1,15 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 8623,
|
||||
"bytes": 8721,
|
||||
"imports": []
|
||||
},
|
||||
"package.json": {
|
||||
"bytes": 3554,
|
||||
"bytes": 3616,
|
||||
"imports": []
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytes": 1912,
|
||||
"bytes": 1941,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -133,7 +133,7 @@
|
|||
]
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytes": 3111,
|
||||
"bytes": 2105,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -144,7 +144,7 @@
|
|||
]
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2928,
|
||||
"bytes": 2979,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -192,7 +192,7 @@
|
|||
]
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytes": 13680,
|
||||
"bytes": 13762,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -213,7 +213,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytes": 3209,
|
||||
"bytes": 3382,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -286,7 +286,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 15788,
|
||||
"bytes": 16020,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -357,7 +357,7 @@
|
|||
"dist/human.esm-nobundle.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 786297
|
||||
"bytes": 785345
|
||||
},
|
||||
"dist/human.esm-nobundle.js": {
|
||||
"imports": [],
|
||||
|
|
@ -375,7 +375,7 @@
|
|||
"bytesInOutput": 51519
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytesInOutput": 12370
|
||||
"bytesInOutput": 12453
|
||||
},
|
||||
"src/face/facemesh.js": {
|
||||
"bytesInOutput": 2461
|
||||
|
|
@ -384,16 +384,16 @@
|
|||
"bytesInOutput": 1118
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytesInOutput": 1830
|
||||
"bytesInOutput": 1563
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytesInOutput": 3000
|
||||
"bytesInOutput": 2736
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 2701
|
||||
"bytesInOutput": 2438
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytesInOutput": 2487
|
||||
"bytesInOutput": 1769
|
||||
},
|
||||
"src/body/modelBase.js": {
|
||||
"bytesInOutput": 866
|
||||
|
|
@ -465,7 +465,7 @@
|
|||
"bytesInOutput": 186
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 2048
|
||||
"bytesInOutput": 2072
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 0
|
||||
|
|
@ -474,13 +474,13 @@
|
|||
"bytesInOutput": 23
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 13232
|
||||
"bytesInOutput": 13473
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 0
|
||||
}
|
||||
},
|
||||
"bytes": 432856
|
||||
"bytes": 431692
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -12316,7 +12316,7 @@
|
|||
]
|
||||
},
|
||||
"package.json": {
|
||||
"bytes": 3554,
|
||||
"bytes": 3616,
|
||||
"imports": []
|
||||
},
|
||||
"src/age/age.js": {
|
||||
|
|
|
|||
|
|
@ -69844,15 +69844,15 @@ return a / b;`;
|
|||
rotatedImage = tf.image.rotateWithOffset(input, angle, 0, faceCenterNormalized);
|
||||
rotationMatrix = util30.buildRotationMatrix(-angle, faceCenter);
|
||||
}
|
||||
const boxCPU = {startPoint: box.startPoint, endPoint: box.endPoint};
|
||||
const face = bounding.cutBoxFromImageAndResize(boxCPU, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const face = bounding.cutBoxFromImageAndResize({startPoint: box.startPoint, endPoint: box.endPoint}, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const outputFace = config.detector.rotation ? tf.image.rotateWithOffset(face, angle) : face;
|
||||
if (!config.mesh.enabled) {
|
||||
const prediction2 = {
|
||||
coords: null,
|
||||
box,
|
||||
faceConfidence: null,
|
||||
confidence: box.confidence,
|
||||
image: face
|
||||
image: outputFace
|
||||
};
|
||||
return prediction2;
|
||||
}
|
||||
|
|
@ -69897,7 +69897,7 @@ return a / b;`;
|
|||
box: landmarksBox,
|
||||
faceConfidence: confidenceVal,
|
||||
confidence: box.confidence,
|
||||
image: face
|
||||
image: outputFace
|
||||
};
|
||||
this.storedBoxes[i] = {...landmarksBox, landmarks: transformedCoords.arraySync(), confidence: box.confidence, faceConfidence: confidenceVal};
|
||||
return prediction;
|
||||
|
|
@ -70009,7 +70009,6 @@ return a / b;`;
|
|||
const models = {};
|
||||
let last = {age: 0};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
async function load(config) {
|
||||
if (!models.age) {
|
||||
models.age = await loadGraphModel(config.face.age.modelPath);
|
||||
|
|
@ -70018,19 +70017,15 @@ return a / b;`;
|
|||
return models.age;
|
||||
}
|
||||
async function predict(image, config) {
|
||||
if (!models.age)
|
||||
return null;
|
||||
if (frame < config.face.age.skipFrames && last.age && last.age > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image.shape[1] * zoom[0] / image.shape[1],
|
||||
image.shape[2] * zoom[1] / image.shape[2],
|
||||
(image.shape[1] - image.shape[1] * zoom[0]) / image.shape[1],
|
||||
(image.shape[2] - image.shape[2] * zoom[1]) / image.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image, box, [0], [config.face.age.inputSize, 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]);
|
||||
tf.dispose(resize);
|
||||
let ageT;
|
||||
|
|
@ -70065,7 +70060,6 @@ return a / b;`;
|
|||
let last = {gender: ""};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
let alternative = false;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load(config) {
|
||||
if (!models.gender) {
|
||||
|
|
@ -70076,19 +70070,15 @@ return a / b;`;
|
|||
return models.gender;
|
||||
}
|
||||
async function predict(image, config) {
|
||||
if (!models.gender)
|
||||
return null;
|
||||
if (frame < config.face.gender.skipFrames && last.gender !== "") {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image.shape[1] * zoom[0] / image.shape[1],
|
||||
image.shape[2] * zoom[1] / image.shape[2],
|
||||
(image.shape[1] - image.shape[1] * zoom[0]) / image.shape[1],
|
||||
(image.shape[2] - image.shape[2] * zoom[1]) / image.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image, box, [0], [config.face.gender.inputSize, config.face.gender.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.gender.inputSize, config.face.gender.inputSize], false);
|
||||
let enhance;
|
||||
if (alternative) {
|
||||
enhance = tf.tidy(() => {
|
||||
|
|
@ -70147,7 +70137,6 @@ return a / b;`;
|
|||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
const scale = 1;
|
||||
async function load(config) {
|
||||
|
|
@ -70158,19 +70147,15 @@ return a / b;`;
|
|||
return models.emotion;
|
||||
}
|
||||
async function predict(image, config) {
|
||||
if (!models.emotion)
|
||||
return null;
|
||||
if (frame < config.face.emotion.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image.shape[1] * zoom[0] / image.shape[1],
|
||||
image.shape[2] * zoom[1] / image.shape[2],
|
||||
(image.shape[1] - image.shape[1] * zoom[0]) / image.shape[1],
|
||||
(image.shape[2] - image.shape[2] * zoom[1]) / image.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image, box, [0], [config.face.emotion.inputSize, 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();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
|
|
@ -70217,10 +70202,6 @@ return a / b;`;
|
|||
var require_embedding = __commonJS((exports) => {
|
||||
const profile = __toModule(require_profile());
|
||||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load(config) {
|
||||
if (!models.embedding) {
|
||||
models.embedding = await loadGraphModel(config.face.embedding.modelPath);
|
||||
|
|
@ -70228,57 +70209,37 @@ return a / b;`;
|
|||
}
|
||||
return models.embedding;
|
||||
}
|
||||
async function predict(image, config) {
|
||||
if (frame < config.face.embedding.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
function simmilarity(embedding1, embedding2) {
|
||||
if ((embedding1 == null ? void 0 : embedding1.length) !== (embedding2 == null ? void 0 : embedding2.length))
|
||||
return 0;
|
||||
const distance = 10 * Math.sqrt(embedding1.map((val, i) => val - embedding2[i]).reduce((dist2, diff) => dist2 + diff ** 2, 0));
|
||||
const confidence = 2 * (0.5 - distance);
|
||||
return Math.trunc(1e3 * confidence) / 1e3;
|
||||
}
|
||||
frame = 0;
|
||||
async function predict(image, config) {
|
||||
if (!models.embedding)
|
||||
return null;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image.shape[1] * zoom[0] / image.shape[1],
|
||||
image.shape[2] * zoom[1] / image.shape[2],
|
||||
(image.shape[1] - image.shape[1] * zoom[0]) / image.shape[1],
|
||||
(image.shape[2] - image.shape[2] * zoom[1]) / image.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image, box, [0], [config.face.embedding.inputSize, config.face.embedding.inputSize]);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
const greenNorm = tf.mul(green, rgb[1]);
|
||||
const blueNorm = tf.mul(blue, rgb[2]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const normalize = tf.tidy(() => grayscale.sub(0.5).mul(2));
|
||||
grayscale.dispose();
|
||||
const obj = [];
|
||||
const resize = tf.image.resizeBilinear(image, [config.face.embedding.inputSize, config.face.embedding.inputSize], false);
|
||||
let data = [];
|
||||
if (config.face.embedding.enabled) {
|
||||
let data;
|
||||
if (!config.profile) {
|
||||
console.log("model", models.embedding);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: normalize});
|
||||
data = embeddingT.dataSync();
|
||||
console.log("embedding", data);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: resize});
|
||||
data = [...embeddingT.dataSync()];
|
||||
tf.dispose(embeddingT);
|
||||
} else {
|
||||
const profileData = await tf.profile(() => models.embedding.predict(normalize));
|
||||
data = profileData.result.dataSync();
|
||||
const profileData = await tf.profile(() => models.embedding.predict({img_inputs: resize}));
|
||||
data = [...profileData.result.dataSync()];
|
||||
profileData.result.dispose();
|
||||
profile.run("emotion", profileData);
|
||||
}
|
||||
obj.sort((a, b) => b.score - a.score);
|
||||
}
|
||||
normalize.dispose();
|
||||
last = obj;
|
||||
resolve(obj);
|
||||
resize.dispose();
|
||||
resolve(data);
|
||||
});
|
||||
}
|
||||
exports.predict = predict;
|
||||
exports.simmilarity = simmilarity;
|
||||
exports.load = load;
|
||||
});
|
||||
|
||||
|
|
@ -89818,6 +89779,11 @@ return a / b;`;
|
|||
}
|
||||
return null;
|
||||
}
|
||||
simmilarity(embedding1, embedding2) {
|
||||
if (this.config.face.embedding.enabled)
|
||||
return embedding.simmilarity(embedding1, embedding2);
|
||||
return 0;
|
||||
}
|
||||
async load(userConfig) {
|
||||
this.state = "load";
|
||||
const timeStamp = now2();
|
||||
|
|
@ -89991,6 +89957,7 @@ return a / b;`;
|
|||
return process3.canvas;
|
||||
}
|
||||
async detect(input, userConfig = {}) {
|
||||
return new Promise(async (resolve) => {
|
||||
this.state = "config";
|
||||
let timeStamp;
|
||||
this.config = mergeDeep(this.config, userConfig);
|
||||
|
|
@ -90000,9 +89967,8 @@ return a / b;`;
|
|||
const error = this.sanity(input);
|
||||
if (error) {
|
||||
this.log(error, input);
|
||||
return {error};
|
||||
resolve({error});
|
||||
}
|
||||
return new Promise(async (resolve) => {
|
||||
let poseRes;
|
||||
let handRes;
|
||||
let faceRes;
|
||||
|
|
@ -90070,10 +90036,12 @@ return a / b;`;
|
|||
resolve({face: faceRes, body: poseRes, hand: handRes, gesture: gestureRes, performance: this.perf, canvas: process3.canvas});
|
||||
});
|
||||
}
|
||||
async warmup(userConfig) {
|
||||
const warmup = new ImageData(255, 255);
|
||||
await this.detect(warmup, userConfig);
|
||||
async warmup(userConfig, sample) {
|
||||
if (!sample)
|
||||
sample = new ImageData(255, 255);
|
||||
const warmup = await this.detect(sample, userConfig);
|
||||
this.log("warmed up");
|
||||
return warmup;
|
||||
}
|
||||
}
|
||||
});
|
||||
|
|
@ -99203,6 +99171,7 @@ return a / b;`;
|
|||
detector: {
|
||||
modelPath: "../models/blazeface-back.json",
|
||||
inputSize: 256,
|
||||
rotation: false,
|
||||
maxFaces: 10,
|
||||
skipFrames: 15,
|
||||
minConfidence: 0.5,
|
||||
|
|
@ -99240,7 +99209,7 @@ return a / b;`;
|
|||
modelPath: "../models/emotion-large.json"
|
||||
},
|
||||
embedding: {
|
||||
enabled: true,
|
||||
enabled: false,
|
||||
inputSize: 112,
|
||||
modelPath: "../models/mobilefacenet.json"
|
||||
}
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 8623,
|
||||
"bytes": 8721,
|
||||
"imports": []
|
||||
},
|
||||
"node_modules/@tensorflow/tfjs-backend-wasm/dist/backend_wasm.js": {
|
||||
|
|
@ -12316,11 +12316,11 @@
|
|||
]
|
||||
},
|
||||
"package.json": {
|
||||
"bytes": 3554,
|
||||
"bytes": 3616,
|
||||
"imports": []
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytes": 1912,
|
||||
"bytes": 1941,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12444,7 +12444,7 @@
|
|||
]
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytes": 3111,
|
||||
"bytes": 2105,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12455,7 +12455,7 @@
|
|||
]
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2928,
|
||||
"bytes": 2979,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12503,7 +12503,7 @@
|
|||
]
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytes": 13680,
|
||||
"bytes": 13762,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12524,7 +12524,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytes": 3209,
|
||||
"bytes": 3382,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12597,7 +12597,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 15788,
|
||||
"bytes": 16020,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -12695,7 +12695,7 @@
|
|||
"dist/human.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 5451131
|
||||
"bytes": 5449697
|
||||
},
|
||||
"dist/human.js": {
|
||||
"imports": [],
|
||||
|
|
@ -12740,7 +12740,7 @@
|
|||
"bytesInOutput": 59051
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytesInOutput": 12860
|
||||
"bytesInOutput": 12942
|
||||
},
|
||||
"src/face/facemesh.js": {
|
||||
"bytesInOutput": 2556
|
||||
|
|
@ -12749,16 +12749,16 @@
|
|||
"bytesInOutput": 1156
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytesInOutput": 1906
|
||||
"bytesInOutput": 1639
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytesInOutput": 3123
|
||||
"bytesInOutput": 2859
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 2809
|
||||
"bytesInOutput": 2546
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytesInOutput": 2590
|
||||
"bytesInOutput": 1833
|
||||
},
|
||||
"src/body/modelBase.js": {
|
||||
"bytesInOutput": 920
|
||||
|
|
@ -12815,7 +12815,7 @@
|
|||
"bytesInOutput": 5588
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 15261
|
||||
"bytesInOutput": 15530
|
||||
},
|
||||
"src/tf.js": {
|
||||
"bytesInOutput": 46
|
||||
|
|
@ -13418,7 +13418,7 @@
|
|||
"bytesInOutput": 3038
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 2254
|
||||
"bytesInOutput": 2280
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 0
|
||||
|
|
@ -13427,7 +13427,7 @@
|
|||
"bytesInOutput": 26
|
||||
}
|
||||
},
|
||||
"bytes": 3647380
|
||||
"bytes": 3646206
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -4275,15 +4275,15 @@ var require_facepipeline = __commonJS((exports2) => {
|
|||
rotatedImage = tf.image.rotateWithOffset(input, angle, 0, faceCenterNormalized);
|
||||
rotationMatrix = util.buildRotationMatrix(-angle, faceCenter);
|
||||
}
|
||||
const boxCPU = {startPoint: box.startPoint, endPoint: box.endPoint};
|
||||
const face2 = bounding.cutBoxFromImageAndResize(boxCPU, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const face2 = bounding.cutBoxFromImageAndResize({startPoint: box.startPoint, endPoint: box.endPoint}, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const outputFace = config2.detector.rotation ? tf.image.rotateWithOffset(face2, angle) : face2;
|
||||
if (!config2.mesh.enabled) {
|
||||
const prediction2 = {
|
||||
coords: null,
|
||||
box,
|
||||
faceConfidence: null,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
return prediction2;
|
||||
}
|
||||
|
|
@ -4328,7 +4328,7 @@ var require_facepipeline = __commonJS((exports2) => {
|
|||
box: landmarksBox,
|
||||
faceConfidence: confidenceVal,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
this.storedBoxes[i] = {...landmarksBox, landmarks: transformedCoords.arraySync(), confidence: box.confidence, faceConfidence: confidenceVal};
|
||||
return prediction;
|
||||
|
|
@ -4440,7 +4440,6 @@ var require_age = __commonJS((exports2) => {
|
|||
const models = {};
|
||||
let last = {age: 0};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
async function load2(config2) {
|
||||
if (!models.age) {
|
||||
models.age = await loadGraphModel(config2.face.age.modelPath);
|
||||
|
|
@ -4449,19 +4448,15 @@ var require_age = __commonJS((exports2) => {
|
|||
return models.age;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.age)
|
||||
return null;
|
||||
if (frame < config2.face.age.skipFrames && last.age && last.age > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.age.inputSize, config2.face.age.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.age.inputSize, config2.face.age.inputSize], false);
|
||||
const enhance = tf.mul(resize, [255]);
|
||||
tf.dispose(resize);
|
||||
let ageT;
|
||||
|
|
@ -4496,7 +4491,6 @@ var require_gender = __commonJS((exports2) => {
|
|||
let last = {gender: ""};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
let alternative = false;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.gender) {
|
||||
|
|
@ -4507,19 +4501,15 @@ var require_gender = __commonJS((exports2) => {
|
|||
return models.gender;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.gender)
|
||||
return null;
|
||||
if (frame < config2.face.gender.skipFrames && last.gender !== "") {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.gender.inputSize, config2.face.gender.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.gender.inputSize, config2.face.gender.inputSize], false);
|
||||
let enhance;
|
||||
if (alternative) {
|
||||
enhance = tf.tidy(() => {
|
||||
|
|
@ -4578,7 +4568,6 @@ var require_emotion = __commonJS((exports2) => {
|
|||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
const scale = 1;
|
||||
async function load2(config2) {
|
||||
|
|
@ -4589,19 +4578,15 @@ var require_emotion = __commonJS((exports2) => {
|
|||
return models.emotion;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.emotion)
|
||||
return null;
|
||||
if (frame < config2.face.emotion.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.emotion.inputSize, config2.face.emotion.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.emotion.inputSize, config2.face.emotion.inputSize], false);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
|
|
@ -4648,10 +4633,6 @@ var require_emotion = __commonJS((exports2) => {
|
|||
var require_embedding = __commonJS((exports2) => {
|
||||
const profile2 = __toModule(require_profile());
|
||||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.embedding) {
|
||||
models.embedding = await loadGraphModel(config2.face.embedding.modelPath);
|
||||
|
|
@ -4659,57 +4640,37 @@ var require_embedding = __commonJS((exports2) => {
|
|||
}
|
||||
return models.embedding;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (frame < config2.face.embedding.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
function simmilarity2(embedding1, embedding2) {
|
||||
if ((embedding1 == null ? void 0 : embedding1.length) !== (embedding2 == null ? void 0 : embedding2.length))
|
||||
return 0;
|
||||
const distance = 10 * Math.sqrt(embedding1.map((val, i) => val - embedding2[i]).reduce((dist2, diff) => dist2 + diff ** 2, 0));
|
||||
const confidence = 2 * (0.5 - distance);
|
||||
return Math.trunc(1e3 * confidence) / 1e3;
|
||||
}
|
||||
frame = 0;
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.embedding)
|
||||
return null;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.embedding.inputSize, config2.face.embedding.inputSize]);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
const greenNorm = tf.mul(green, rgb[1]);
|
||||
const blueNorm = tf.mul(blue, rgb[2]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const normalize = tf.tidy(() => grayscale.sub(0.5).mul(2));
|
||||
grayscale.dispose();
|
||||
const obj = [];
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.embedding.inputSize, config2.face.embedding.inputSize], false);
|
||||
let data2 = [];
|
||||
if (config2.face.embedding.enabled) {
|
||||
let data2;
|
||||
if (!config2.profile) {
|
||||
console.log("model", models.embedding);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: normalize});
|
||||
data2 = embeddingT.dataSync();
|
||||
console.log("embedding", data2);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: resize});
|
||||
data2 = [...embeddingT.dataSync()];
|
||||
tf.dispose(embeddingT);
|
||||
} else {
|
||||
const profileData = await tf.profile(() => models.embedding.predict(normalize));
|
||||
data2 = profileData.result.dataSync();
|
||||
const profileData = await tf.profile(() => models.embedding.predict({img_inputs: resize}));
|
||||
data2 = [...profileData.result.dataSync()];
|
||||
profileData.result.dispose();
|
||||
profile2.run("emotion", profileData);
|
||||
}
|
||||
obj.sort((a, b) => b.score - a.score);
|
||||
}
|
||||
normalize.dispose();
|
||||
last = obj;
|
||||
resolve(obj);
|
||||
resize.dispose();
|
||||
resolve(data2);
|
||||
});
|
||||
}
|
||||
exports2.predict = predict2;
|
||||
exports2.simmilarity = simmilarity2;
|
||||
exports2.load = load2;
|
||||
});
|
||||
|
||||
|
|
@ -24363,6 +24324,7 @@ var config_default = {
|
|||
detector: {
|
||||
modelPath: "../models/blazeface-back.json",
|
||||
inputSize: 256,
|
||||
rotation: false,
|
||||
maxFaces: 10,
|
||||
skipFrames: 15,
|
||||
minConfidence: 0.5,
|
||||
|
|
@ -24400,7 +24362,7 @@ var config_default = {
|
|||
modelPath: "../models/emotion-large.json"
|
||||
},
|
||||
embedding: {
|
||||
enabled: true,
|
||||
enabled: false,
|
||||
inputSize: 112,
|
||||
modelPath: "../models/mobilefacenet.json"
|
||||
}
|
||||
|
|
@ -24523,6 +24485,11 @@ class Human {
|
|||
}
|
||||
return null;
|
||||
}
|
||||
simmilarity(embedding1, embedding2) {
|
||||
if (this.config.face.embedding.enabled)
|
||||
return embedding.simmilarity(embedding1, embedding2);
|
||||
return 0;
|
||||
}
|
||||
async load(userConfig) {
|
||||
this.state = "load";
|
||||
const timeStamp = now();
|
||||
|
|
@ -24696,6 +24663,7 @@ class Human {
|
|||
return process3.canvas;
|
||||
}
|
||||
async detect(input, userConfig = {}) {
|
||||
return new Promise(async (resolve) => {
|
||||
this.state = "config";
|
||||
let timeStamp;
|
||||
this.config = mergeDeep(this.config, userConfig);
|
||||
|
|
@ -24705,9 +24673,8 @@ class Human {
|
|||
const error = this.sanity(input);
|
||||
if (error) {
|
||||
this.log(error, input);
|
||||
return {error};
|
||||
resolve({error});
|
||||
}
|
||||
return new Promise(async (resolve) => {
|
||||
let poseRes;
|
||||
let handRes;
|
||||
let faceRes;
|
||||
|
|
@ -24775,10 +24742,12 @@ class Human {
|
|||
resolve({face: faceRes, body: poseRes, hand: handRes, gesture: gestureRes, performance: this.perf, canvas: process3.canvas});
|
||||
});
|
||||
}
|
||||
async warmup(userConfig) {
|
||||
const warmup = new ImageData(255, 255);
|
||||
await this.detect(warmup, userConfig);
|
||||
async warmup(userConfig, sample) {
|
||||
if (!sample)
|
||||
sample = new ImageData(255, 255);
|
||||
const warmup = await this.detect(sample, userConfig);
|
||||
this.log("warmed up");
|
||||
return warmup;
|
||||
}
|
||||
}
|
||||
//# sourceMappingURL=human.node-nobundle.js.map
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
|
|
@ -95993,15 +95993,15 @@ var require_facepipeline = __commonJS((exports2) => {
|
|||
rotatedImage = tf.image.rotateWithOffset(input, angle, 0, faceCenterNormalized);
|
||||
rotationMatrix = util27.buildRotationMatrix(-angle, faceCenter);
|
||||
}
|
||||
const boxCPU = {startPoint: box.startPoint, endPoint: box.endPoint};
|
||||
const face2 = bounding.cutBoxFromImageAndResize(boxCPU, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const face2 = bounding.cutBoxFromImageAndResize({startPoint: box.startPoint, endPoint: box.endPoint}, rotatedImage, [this.meshHeight, this.meshWidth]).div(255);
|
||||
const outputFace = config2.detector.rotation ? tf.image.rotateWithOffset(face2, angle) : face2;
|
||||
if (!config2.mesh.enabled) {
|
||||
const prediction2 = {
|
||||
coords: null,
|
||||
box,
|
||||
faceConfidence: null,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
return prediction2;
|
||||
}
|
||||
|
|
@ -96046,7 +96046,7 @@ var require_facepipeline = __commonJS((exports2) => {
|
|||
box: landmarksBox,
|
||||
faceConfidence: confidenceVal,
|
||||
confidence: box.confidence,
|
||||
image: face2
|
||||
image: outputFace
|
||||
};
|
||||
this.storedBoxes[i] = {...landmarksBox, landmarks: transformedCoords.arraySync(), confidence: box.confidence, faceConfidence: confidenceVal};
|
||||
return prediction;
|
||||
|
|
@ -96158,7 +96158,6 @@ var require_age = __commonJS((exports2) => {
|
|||
const models = {};
|
||||
let last = {age: 0};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
async function load2(config2) {
|
||||
if (!models.age) {
|
||||
models.age = await loadGraphModel(config2.face.age.modelPath);
|
||||
|
|
@ -96167,19 +96166,15 @@ var require_age = __commonJS((exports2) => {
|
|||
return models.age;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.age)
|
||||
return null;
|
||||
if (frame < config2.face.age.skipFrames && last.age && last.age > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.age.inputSize, config2.face.age.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.age.inputSize, config2.face.age.inputSize], false);
|
||||
const enhance = tf.mul(resize, [255]);
|
||||
tf.dispose(resize);
|
||||
let ageT;
|
||||
|
|
@ -96214,7 +96209,6 @@ var require_gender = __commonJS((exports2) => {
|
|||
let last = {gender: ""};
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
let alternative = false;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.gender) {
|
||||
|
|
@ -96225,19 +96219,15 @@ var require_gender = __commonJS((exports2) => {
|
|||
return models.gender;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.gender)
|
||||
return null;
|
||||
if (frame < config2.face.gender.skipFrames && last.gender !== "") {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.gender.inputSize, config2.face.gender.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.gender.inputSize, config2.face.gender.inputSize], false);
|
||||
let enhance;
|
||||
if (alternative) {
|
||||
enhance = tf.tidy(() => {
|
||||
|
|
@ -96296,7 +96286,6 @@ var require_emotion = __commonJS((exports2) => {
|
|||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
const scale = 1;
|
||||
async function load2(config2) {
|
||||
|
|
@ -96307,19 +96296,15 @@ var require_emotion = __commonJS((exports2) => {
|
|||
return models.emotion;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.emotion)
|
||||
return null;
|
||||
if (frame < config2.face.emotion.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
}
|
||||
frame = 0;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.emotion.inputSize, config2.face.emotion.inputSize]);
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.emotion.inputSize, config2.face.emotion.inputSize], false);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
|
|
@ -96366,10 +96351,6 @@ var require_emotion = __commonJS((exports2) => {
|
|||
var require_embedding = __commonJS((exports2) => {
|
||||
const profile2 = __toModule(require_profile());
|
||||
const models = {};
|
||||
let last = [];
|
||||
let frame = Number.MAX_SAFE_INTEGER;
|
||||
const zoom = [0, 0];
|
||||
const rgb = [0.2989, 0.587, 0.114];
|
||||
async function load2(config2) {
|
||||
if (!models.embedding) {
|
||||
models.embedding = await loadGraphModel(config2.face.embedding.modelPath);
|
||||
|
|
@ -96377,57 +96358,37 @@ var require_embedding = __commonJS((exports2) => {
|
|||
}
|
||||
return models.embedding;
|
||||
}
|
||||
async function predict2(image2, config2) {
|
||||
if (frame < config2.face.embedding.skipFrames && last.length > 0) {
|
||||
frame += 1;
|
||||
return last;
|
||||
function simmilarity2(embedding1, embedding2) {
|
||||
if ((embedding1 == null ? void 0 : embedding1.length) !== (embedding2 == null ? void 0 : embedding2.length))
|
||||
return 0;
|
||||
const distance = 10 * Math.sqrt(embedding1.map((val, i) => val - embedding2[i]).reduce((dist2, diff) => dist2 + diff ** 2, 0));
|
||||
const confidence = 2 * (0.5 - distance);
|
||||
return Math.trunc(1e3 * confidence) / 1e3;
|
||||
}
|
||||
frame = 0;
|
||||
async function predict2(image2, config2) {
|
||||
if (!models.embedding)
|
||||
return null;
|
||||
return new Promise(async (resolve) => {
|
||||
const box = [[
|
||||
image2.shape[1] * zoom[0] / image2.shape[1],
|
||||
image2.shape[2] * zoom[1] / image2.shape[2],
|
||||
(image2.shape[1] - image2.shape[1] * zoom[0]) / image2.shape[1],
|
||||
(image2.shape[2] - image2.shape[2] * zoom[1]) / image2.shape[2]
|
||||
]];
|
||||
const resize = tf.image.cropAndResize(image2, box, [0], [config2.face.embedding.inputSize, config2.face.embedding.inputSize]);
|
||||
const [red, green, blue] = tf.split(resize, 3, 3);
|
||||
resize.dispose();
|
||||
const redNorm = tf.mul(red, rgb[0]);
|
||||
const greenNorm = tf.mul(green, rgb[1]);
|
||||
const blueNorm = tf.mul(blue, rgb[2]);
|
||||
red.dispose();
|
||||
green.dispose();
|
||||
blue.dispose();
|
||||
const grayscale = tf.addN([redNorm, greenNorm, blueNorm]);
|
||||
redNorm.dispose();
|
||||
greenNorm.dispose();
|
||||
blueNorm.dispose();
|
||||
const normalize = tf.tidy(() => grayscale.sub(0.5).mul(2));
|
||||
grayscale.dispose();
|
||||
const obj = [];
|
||||
const resize = tf.image.resizeBilinear(image2, [config2.face.embedding.inputSize, config2.face.embedding.inputSize], false);
|
||||
let data2 = [];
|
||||
if (config2.face.embedding.enabled) {
|
||||
let data2;
|
||||
if (!config2.profile) {
|
||||
console.log("model", models.embedding);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: normalize});
|
||||
data2 = embeddingT.dataSync();
|
||||
console.log("embedding", data2);
|
||||
const embeddingT = await models.embedding.predict({img_inputs: resize});
|
||||
data2 = [...embeddingT.dataSync()];
|
||||
tf.dispose(embeddingT);
|
||||
} else {
|
||||
const profileData = await tf.profile(() => models.embedding.predict(normalize));
|
||||
data2 = profileData.result.dataSync();
|
||||
const profileData = await tf.profile(() => models.embedding.predict({img_inputs: resize}));
|
||||
data2 = [...profileData.result.dataSync()];
|
||||
profileData.result.dispose();
|
||||
profile2.run("emotion", profileData);
|
||||
}
|
||||
obj.sort((a, b) => b.score - a.score);
|
||||
}
|
||||
normalize.dispose();
|
||||
last = obj;
|
||||
resolve(obj);
|
||||
resize.dispose();
|
||||
resolve(data2);
|
||||
});
|
||||
}
|
||||
exports2.predict = predict2;
|
||||
exports2.simmilarity = simmilarity2;
|
||||
exports2.load = load2;
|
||||
});
|
||||
|
||||
|
|
@ -120187,6 +120148,7 @@ var config_default = {
|
|||
detector: {
|
||||
modelPath: "../models/blazeface-back.json",
|
||||
inputSize: 256,
|
||||
rotation: false,
|
||||
maxFaces: 10,
|
||||
skipFrames: 15,
|
||||
minConfidence: 0.5,
|
||||
|
|
@ -120224,7 +120186,7 @@ var config_default = {
|
|||
modelPath: "../models/emotion-large.json"
|
||||
},
|
||||
embedding: {
|
||||
enabled: true,
|
||||
enabled: false,
|
||||
inputSize: 112,
|
||||
modelPath: "../models/mobilefacenet.json"
|
||||
}
|
||||
|
|
@ -120347,6 +120309,11 @@ class Human {
|
|||
}
|
||||
return null;
|
||||
}
|
||||
simmilarity(embedding1, embedding2) {
|
||||
if (this.config.face.embedding.enabled)
|
||||
return embedding.simmilarity(embedding1, embedding2);
|
||||
return 0;
|
||||
}
|
||||
async load(userConfig) {
|
||||
this.state = "load";
|
||||
const timeStamp = now();
|
||||
|
|
@ -120520,6 +120487,7 @@ class Human {
|
|||
return process3.canvas;
|
||||
}
|
||||
async detect(input, userConfig = {}) {
|
||||
return new Promise(async (resolve) => {
|
||||
this.state = "config";
|
||||
let timeStamp;
|
||||
this.config = mergeDeep(this.config, userConfig);
|
||||
|
|
@ -120529,9 +120497,8 @@ class Human {
|
|||
const error = this.sanity(input);
|
||||
if (error) {
|
||||
this.log(error, input);
|
||||
return {error};
|
||||
resolve({error});
|
||||
}
|
||||
return new Promise(async (resolve) => {
|
||||
let poseRes;
|
||||
let handRes;
|
||||
let faceRes;
|
||||
|
|
@ -120599,10 +120566,12 @@ class Human {
|
|||
resolve({face: faceRes, body: poseRes, hand: handRes, gesture: gestureRes, performance: this.perf, canvas: process3.canvas});
|
||||
});
|
||||
}
|
||||
async warmup(userConfig) {
|
||||
const warmup = new ImageData(255, 255);
|
||||
await this.detect(warmup, userConfig);
|
||||
async warmup(userConfig, sample) {
|
||||
if (!sample)
|
||||
sample = new ImageData(255, 255);
|
||||
const warmup = await this.detect(sample, userConfig);
|
||||
this.log("warmed up");
|
||||
return warmup;
|
||||
}
|
||||
}
|
||||
//# sourceMappingURL=human.node.js.map
|
||||
|
|
|
|||
File diff suppressed because one or more lines are too long
|
|
@ -1,15 +1,15 @@
|
|||
{
|
||||
"inputs": {
|
||||
"config.js": {
|
||||
"bytes": 8623,
|
||||
"bytes": 8721,
|
||||
"imports": []
|
||||
},
|
||||
"package.json": {
|
||||
"bytes": 3554,
|
||||
"bytes": 3616,
|
||||
"imports": []
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytes": 1912,
|
||||
"bytes": 1941,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -133,7 +133,7 @@
|
|||
]
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytes": 3111,
|
||||
"bytes": 2105,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -144,7 +144,7 @@
|
|||
]
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytes": 2928,
|
||||
"bytes": 2979,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -192,7 +192,7 @@
|
|||
]
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytes": 13680,
|
||||
"bytes": 13762,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -213,7 +213,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytes": 3209,
|
||||
"bytes": 3382,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -286,7 +286,7 @@
|
|||
"imports": []
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytes": 15788,
|
||||
"bytes": 16020,
|
||||
"imports": [
|
||||
{
|
||||
"path": "src/tf.js"
|
||||
|
|
@ -357,7 +357,7 @@
|
|||
"dist/human.node-nobundle.js.map": {
|
||||
"imports": [],
|
||||
"inputs": {},
|
||||
"bytes": 802566
|
||||
"bytes": 801852
|
||||
},
|
||||
"dist/human.node-nobundle.js": {
|
||||
"imports": [],
|
||||
|
|
@ -375,7 +375,7 @@
|
|||
"bytesInOutput": 51530
|
||||
},
|
||||
"src/face/facepipeline.js": {
|
||||
"bytesInOutput": 12372
|
||||
"bytesInOutput": 12455
|
||||
},
|
||||
"src/face/facemesh.js": {
|
||||
"bytesInOutput": 2465
|
||||
|
|
@ -384,16 +384,16 @@
|
|||
"bytesInOutput": 1120
|
||||
},
|
||||
"src/age/age.js": {
|
||||
"bytesInOutput": 1833
|
||||
"bytesInOutput": 1566
|
||||
},
|
||||
"src/gender/gender.js": {
|
||||
"bytesInOutput": 3003
|
||||
"bytesInOutput": 2739
|
||||
},
|
||||
"src/emotion/emotion.js": {
|
||||
"bytesInOutput": 2704
|
||||
"bytesInOutput": 2441
|
||||
},
|
||||
"src/embedding/embedding.js": {
|
||||
"bytesInOutput": 2490
|
||||
"bytesInOutput": 1773
|
||||
},
|
||||
"src/body/modelBase.js": {
|
||||
"bytesInOutput": 868
|
||||
|
|
@ -468,7 +468,7 @@
|
|||
"bytesInOutput": 186
|
||||
},
|
||||
"config.js": {
|
||||
"bytesInOutput": 2048
|
||||
"bytesInOutput": 2072
|
||||
},
|
||||
"package.json": {
|
||||
"bytesInOutput": 0
|
||||
|
|
@ -477,10 +477,10 @@
|
|||
"bytesInOutput": 23
|
||||
},
|
||||
"src/human.js": {
|
||||
"bytesInOutput": 13237
|
||||
"bytesInOutput": 13478
|
||||
}
|
||||
},
|
||||
"bytes": 433211
|
||||
"bytes": 432048
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -2306,9 +2306,9 @@
|
|||
"dev": true
|
||||
},
|
||||
"simple-git": {
|
||||
"version": "2.21.0",
|
||||
"resolved": "https://registry.npmjs.org/simple-git/-/simple-git-2.21.0.tgz",
|
||||
"integrity": "sha512-rohCHmEjD/ESXFLxF4bVeqgdb4Awc65ZyyuCKl3f7BvgMbZOBa/Ye3HN/GFnvruiUOAWWNupxhz3Rz5/3vJLTg==",
|
||||
"version": "2.22.0",
|
||||
"resolved": "https://registry.npmjs.org/simple-git/-/simple-git-2.22.0.tgz",
|
||||
"integrity": "sha512-/8WMNrQ5slYl05jYOpeh4BIyMQc84PkEvk9YAcBHVJaPoAgoxzLzdRzUzWDZJ9U6Z2pUjLxkGcQC0kU4pbRN5Q==",
|
||||
"dev": true,
|
||||
"requires": {
|
||||
"@kwsites/file-exists": "^1.1.1",
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"name": "@vladmandic/human",
|
||||
"version": "0.8.8",
|
||||
"description": "human: 3D Face Detection, Body Pose, Hand & Finger Tracking, Iris Tracking, Age & Gender Prediction, Emotion Prediction & Gesture Recognition",
|
||||
"description": "Human: 3D Face Detection, Face Embedding, Body Pose, Hand & Finger Tracking, Iris Tracking, Age & Gender Prediction, Emotion Prediction & Gesture Recognition",
|
||||
"sideEffects": false,
|
||||
"main": "dist/human.node.js",
|
||||
"module": "dist/human.esm.js",
|
||||
|
|
@ -37,7 +37,7 @@
|
|||
"eslint-plugin-promise": "^4.2.1",
|
||||
"rimraf": "^3.0.2",
|
||||
"seedrandom": "^3.0.5",
|
||||
"simple-git": "^2.21.0"
|
||||
"simple-git": "^2.22.0"
|
||||
},
|
||||
"scripts": {
|
||||
"start": "node --trace-warnings --unhandled-rejections=strict --trace-uncaught --no-deprecation src/node.js",
|
||||
|
|
@ -57,6 +57,8 @@
|
|||
"tensorflowjs",
|
||||
"face-detection",
|
||||
"face-geometry",
|
||||
"face-embedding",
|
||||
"face-recognition",
|
||||
"body-tracking",
|
||||
"hand-tracking",
|
||||
"iris-tracking",
|
||||
|
|
|
|||
2
wiki
2
wiki
|
|
@ -1 +1 @@
|
|||
Subproject commit 726b24b9a76fd3df36fc309bd4f962daf4c74e45
|
||||
Subproject commit 1a8d273b7fda90f496a792fd2a16c6978a88baa3
|
||||
Loading…
Reference in New Issue