Video-driven Talking Head Generation (Cross-Reenactment) on HDTF benchmark leaderboard · SOTA2 Research

Video-driven Talking Head Generation (Cross-Reenactment) on HDTF

27.85FIDMGGTalk

Evaluation Results

Method	Links
MGGTalk2025.04		27.85	0.223	0.241
Portrait4D-v22025.04		42.82	0.238	0.248
DaGAN2025.04		48.2	0.225	0.258
OTAvatar2025.04		50.37	0.21	0.312
Real3DPortrait2025.04		51.36	0.256	0.293
Styleheat2025.04		76.43	0.369	0.334
ROME2025.04		79.31	0.229	0.286