Skip to content

Latest commit

 

History

History
2 lines (2 loc) · 366 Bytes

File metadata and controls

2 lines (2 loc) · 366 Bytes

Masked-Face-Detection-Reconstruction-For-Identity-Verification

We implemented MTCNN for face extraction, ESRGAN for feature extraction, and Sequential API for gender classification achieving 94.5% accuracy. YOLOv9 for mask segmentation with 92.3% accuracy. Landmark detection via MediaPipe. Pix2Pix for realistic facial feature generation, achieving 82% accuracy.