Core Features:
- On-device face recognition with InsightFace AI models
- Real-time face detection with Buffalo_L model
- 106-point facial landmark extraction
- Face feature extraction and matching
- Anti-spoofing detection support
Platform Support:
- iOS 12.0+ support with static library integration
- Android API 21+ support with JNI libraries
- Cross-platform ONNX model inference
Example App:
- Simple version using Service API (easy integration)
- Advanced version with custom isolate (performance optimization)
- Tab-based UI to switch between Simple and Advanced modes
- Real-time camera preview and face processing
Architecture:
- FFI-based native library integration
- Isolate support for non-blocking AI inference
- Strongly-typed Result models
- Model download manager with cloud storage support
Performance:
- Complete face recognition pipeline < 500ms
- Optimized image conversion and processing
- Automatic resolution adjustment for performance