Huihui Yin

Dual-domain adversarial training for realistic 4D facial makeup detection and generation

The results indicate that developing domain-aware alignment and hybrid loss integration techniques is beneficial for effective facial analysis in both controlled and challenging environments.