# vladmandic/human

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

Repository: https://github.com/vladmandic/human
Canonical: https://ross.abutalabs.com/products/human
Homepage: https://vladmandic.github.io/human/demo/index.html
Language: HTML
License: MIT
License Family: permissive
Topics: face-detection, gender-prediction, iris-tracking, body-tracking, hand-tracking, tensorflowjs, age-estimation, face-position, tfjs, emotion-detection, gesture-recognition, face-recognition, face-mesh, gaze-tracking, body-segmentation, face-matching, faceid
Last push: 2025-12-13T09:49:12+00:00

## Health v2 (maintenance only)
Score: 48/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 57, release rhythm 8, longevity 100
- inputs: {"age_days": 2152, "days_push": 263, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3264, forks 449 (observed 2026-08-28T04:07:52.873148+00:00)

## What it is
Human is a JavaScript/TypeScript library built on TensorFlow.js that combines multiple ML models for 3D face detection and recognition, body pose tracking, 3D hand and finger tracking, iris analysis, and age, gender, and emotion prediction. It runs in browsers (WebGPU, WebGL, WASM, CPU) and Node.js with a unified API, attention pipeline, and built-in image, video, and webcam handling.

## Use cases
- detect and track faces in a webcam stream
- recognize faces and match identities in the browser
- track body pose from video for fitness apps
- track 3D hand and finger movements
- estimate age and gender from photos
- detect emotions from facial expressions
- track gaze and iris position
- recognize hand gestures in real time

## When to choose
- you need multi-model human perception (face, pose, hands, gestures) with a single unified API
- you want real-time inference in the browser without a server backend
- you need both browser and Node.js support with GPU or WASM backends
- you want temporal smoothing and attention pipelines out of the box

## When to avoid
- you need server-side Python or native ML frameworks rather than TensorFlow.js
- you need maximum accuracy over speed for production biometric systems
- you only need a single simple detection task where a lighter model would suffice
- you require Node.js 23.x, which is not supported

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, computer-vision, image-processing, nlp
- domain: computer-vision, machine-learning, artificial-intelligence, web-development
- platform: browser, wasm, cross-platform
- tags: face-detection, face-recognition, pose-tracking, hand-tracking, tensorflowjs, gesture-recognition, age-estimation, emotion-detection, gaze-tracking, body-segmentation, webgpu, webgl, nodejs, web-server

## Member repositories
- vladmandic/human (main) score 48

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:52.873148+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:23:11.351531+00:00, confidence not recorded.
  - readme: https://github.com/vladmandic/human (fetched 2026-08-28T04:07:52.873148+00:00, sha a779149cda82)
  - homepage: https://vladmandic.github.io/human/demo/index.html (fetched 2026-08-29T09:36:33.978377+00:00, sha d64023d5e786)
- Data as of 2026-08-30T08:39:29.467469+00:00.