vladmandic/human
A TensorFlow.js-based computer vision library providing face detection, body pose tracking, hand tracking, and gesture recognition with support for WebGPU, WebGL, WASM, and CPU backends.
Not currently ranked — collecting fresh signals.
star history
The Human library is a browser and Node.js compatible computer vision framework that combines multiple machine learning models for real-time analysis. It performs face detection and recognition, body pose estimation, hand and finger tracking, iris analysis, and predicts age, gender, emotion, and gaze. The system features intelligent frame change detection and temporal interpolation for optimized performance across different hardware backends.
Frequently asked
- What is vladmandic/human?
- A TensorFlow.js-based computer vision library providing face detection, body pose tracking, hand tracking, and gesture recognition with support for WebGPU, WebGL, WASM, and CPU backends.
- Is human open source?
- Yes — vladmandic/human is open source, released under the MIT license.
- What language is human written in?
- vladmandic/human is primarily written in HTML.
- How popular is human?
- vladmandic/human has 3.2k stars on GitHub.
- Where can I find human?
- vladmandic/human is on GitHub at https://github.com/vladmandic/human.