Description
Face features usually mean one library for detection, another for landmarks and another for recognition, each with different formats. UniFace puts them under a single Python API: detection, alignment, landmarks, face mesh, recognition, parsing, gaze, attributes and anti-spoofing.
It runs on ONNX Runtime, is lightweight and production-ready, and one FaceAnalyzer call handles detection, alignment and recognition while attribute models stay opt-in.
Fifteen tasks:Detection, landmarks, recognition, parsing, gaze and more.
One API:Detect, align and recognize in one call.
Opt-in models:Load attribute predictors as needed.
ONNX:Lightweight deployment.
It runs on ONNX Runtime, is lightweight and production-ready, and one FaceAnalyzer call handles detection, alignment and recognition while attribute models stay opt-in.
Features
Fifteen tasks:Detection, landmarks, recognition, parsing, gaze and more.
One API:Detect, align and recognize in one call.
Opt-in models:Load attribute predictors as needed.
ONNX:Lightweight deployment.
