lithiumice/syncnet

Model Checkpoints

Pre-trained Models

This project includes the following pre-trained model weights:

S3FD Face Detection

• Filename: sfd_face.pth
• Purpose: High-accuracy face detection
• Architecture: Single Shot Scale-invariant Face Detector (S3FD)

SyncNet Audio-Visual Synchronization

• Filename: syncnet_v2.model
• Purpose: Audio-visual synchronization detection
• Original Repo: https://github.com/joonson/syncnet_python