face_live 0.2.1

face_live: ^0.2.1 copied to clipboard

🐛 Bug Fixes #

• iOS Recording Error: Fixed AVFoundation error -11803 "Cannot Record" by implementing proper coordination between video recording and photo capture operations
• Enhanced iOS Stability: Added completion flags and delayed session stopping (0.3s) to ensure both video and image operations complete successfully before finalizing

🔧 Improvements #

• Better Error Handling: Improved logging for video recording and photo capture operations on iOS
• Coordination Pattern: Implemented robust completion tracking to prevent partial results and ensure reliable capture on all iOS devices

🎉 New Features #

• Pitch Detection: Added enablePitchDetection parameter to include up/down head movement in liveness progress
• Capture Delay: Added captureDelayMillis parameter for optimal user positioning before final capture

🔧 Improvements #

• Adjusted targetYawSpan default from 100° to 65° for easier and more user-friendly liveness detection
• Enhanced native implementations on both Android and iOS to support new parameters

📱 Platform Support #

• Android: Updated FaceLivenessAnalyzer and CameraView with new parameter support
• iOS: Updated FaceLivenessCameraView with pitch detection and capture delay features

🎉 New Features #

• Image Capture: Added onImageCapture callback for capturing still images alongside video recording
• Enhanced Security: Stricter liveness detection with improved validation parameters

🔧 Improvements #

• Increased target yaw span from 80° to 100° for more comprehensive head movement
• Increased minimum completion time from 3s to 4s to prevent rushed attempts
• Increased minimum face size requirement from 15% to 20% of frame area
• Reduced maximum consecutive missed frames from 10 to 5 for better continuity
• Enhanced bidirectional movement thresholds from ±5° to ±10° for clearer left/right detection
• Switched ML Kit performance mode from FAST to ACCURATE for better face detection
• Updated example app to demonstrate both video and image outputs

⚠️ Breaking Changes #

• onLivenessSuccess now returns a map with videoPath and imagePath instead of a single string
• More stringent liveness validation may require users to be more deliberate in their movements

📱 Platform Support #

• Android: Added CameraX ImageCapture integration
• iOS: Added AVCapturePhotoOutput integration

• Initial release of face_live plugin
• Real-time face liveness detection using ML Kit
• Cross-platform support for Android (API 21+) and iOS (15.5+)
• Native camera preview with high performance
• Head movement tracking based on yaw rotation
• Progress updates (0-100%) during liveness check
• Video recording of liveness session
• Example app demonstrating usage