Human Library

AI-powered 3D Face Detection & Rotation Tracking, Face Description & Recognition,
Body Pose Tracking, 3D Hand & Finger Tracking, Iris Analysis,
Age & Gender & Emotion Prediction, Gesture Recognition

JavaScript module using TensorFlow/JS Machine Learning library

Browser:
Compatible with both desktop and mobile platforms
Compatible with CPU, WebGL, WASM backends
Compatible with WebWorker execution
NodeJS:
Compatible with both software tfjs-node and
GPU accelerated backends tfjs-node-gpu using CUDA libraries

Check out Live Demo for processing of live WebCam video or static images

Demos

Project pages

Wiki pages

Additional notes

See issues and discussions for list of known limitations and planned enhancements

Suggestions are welcome!

Options

As presented in the demo application...

demo/index.html

Examples

Training image:

demo/index.html

Using static images:

demo/index.html

Live WebCam view:

demo/index.html

Face Similarity Matching:

demo/facematch.html

Face3D OpenGL Rendering:

demo/face3d.html

468-Point Face Mesh Defails:
(view in full resolution to see keypoints)

Quick Start

Simply load Human (IIFE version) directly from a cloud CDN in your HTML file:
(pick one: jsdelirv, unpkg or cdnjs)

<script src="https://cdn.jsdelivr.net/npm/@vladmandic/human/dist/human.js"></script>
<script src="https://unpkg.dev/@vladmandic/human/dist/human.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/human/1.4.1/human.js"></script>

For details, including how to use Browser ESM version or NodeJS version of Human, see Installation

Inputs

Human library can process all known input types:

Image, ImageData, ImageBitmap, Canvas, OffscreenCanvas, Tensor,
HTMLImageElement, HTMLCanvasElement, HTMLVideoElement, HTMLMediaElement

Additionally, HTMLVideoElement, HTMLMediaElement can be a standard <video> tag that links to:

WebCam on user's system
Any supported video type
For example: .mp4, .avi, etc.
Additional video types supported via HTML5 Media Source Extensions
Live streaming examples:
- HLS (HTTP Live Streaming) using hls.js
- DASH (Dynamic Adaptive Streaming over HTTP) using dash.js
WebRTC media track

Example

Example simple app that uses Human to process video input and
draw output on screen using internal draw helper functions

// create instance of human with simple configuration using default values
const config = { backend: 'webgl' };
const human = new Human(config);

function detectVideo() {
  // select input HTMLVideoElement and output HTMLCanvasElement from page
  const inputVideo = document.getElementById('video-id');
  const outputCanvas = document.getElementById('canvas-id');
  // perform processing using default configuration
  human.detect(inputVideo).then((result) => {
    // result object will contain detected details
    // as well as the processed canvas itself
    // so lets first draw processed frame on canvas
    human.draw.canvas(result.canvas, outputCanvas);
    // then draw results on the same canvas
    human.draw.face(outputCanvas, result.face);
    human.draw.body(outputCanvas, result.body);
    human.draw.hand(outputCanvas, result.hand);
    human.draw.gesture(outputCanvas, result.gesture);
    // and loop immediate to the next frame
    requestAnimationFrame(detectVideo);
  });
}

detectVideo();

or using async/await:

// create instance of human with simple configuration using default values
const config = { backend: 'webgl' };
const human = new Human(config);

async function detectVideo() {
  const inputVideo = document.getElementById('video-id');
  const outputCanvas = document.getElementById('canvas-id');
  const result = await human.detect(inputVideo);
  human.draw.all(outputCanvas, result);
  requestAnimationFrame(detectVideo);
}

detectVideo();

Default models

Default models in Human library are:

Face Detection: MediaPipe BlazeFace (Back version)
Face Mesh: MediaPipe FaceMesh
Face Description: HSE FaceRes
Face Iris Analysis: MediaPipe Iris
Emotion Detection: Oarriaga Emotion
Body Analysis: PoseNet (AtomicBits version)

Note that alternative models are provided and can be enabled via configuration
For example, PoseNet model can be switched for BlazePose model depending on the use case

For more info, see Configuration Details and List of Models

Human library is written in TypeScript 4.2
Conforming to JavaScript ECMAScript version 2020 standard
Build target is JavaScript EMCAScript version 2018

For details see Wiki Pages
and API Specification