UniFace — комплексный анализ лиц на фото и видео

★ 1K

UniFace is a lightweight, production-ready Python library for comprehensive face analysis: detection, recognition, tracking, landmark analysis, face parsing (segmentation), gaze estimation, and face attributes. Reach for it when your app needs to work with faces uniformly: find faces in photos/video, compare and recognize them, track across frames, and determine traits and gaze — from one library instead of assembling separate models. Its focus is a unified Python toolkit for face-analysis tasks in your projects, not an end application or whole-body pose estimation (OpenPose): the value is covering the main face operations in one package with production in mind. Use it mindful of biometric and privacy laws, people's consent, and purpose (face recognition is heavily regulated). It suits developers embedding face detection/recognition/analysis into their services.