Speech-driven talking face generation on VoxCeleb (test)
6.61LSE-DHRTF
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| HRTFVenue=-2025.10 | 6.61 | 8.65 | 0.11 | 0.67 | 29.66 | 29.28 | 0.0042 | |
| HRTF-SCFPVenue=-2025.10 | 7.08 | 8.23 | 0.12 | 0.65 | 29.42 | 32.66 | 0.0046 | |
| VideoRetalkingVenue=SIGGRAPH 20222025.10 | 9.44 | 3.34 | 0.12 | 0.62 | 29.47 | 39.47 | 0.0061 | |
| AniportaitVenue=ECCV 20242025.10 | 9.63 | 6.39 | 0.13 | 0.56 | 29.54 | 44.13 | 0.0053 | |
| SyncTalkVenue=CVPR 20242025.10 | 11.46 | 4.83 | 0.22 | 0.56 | 29.41 | 46.7 | 0.0064 | |
| HalloVenue=arXiv 2024.72025.10 | 12.43 | 5.14 | 0.24 | 0.59 | 29.42 | 44.82 | 0.0059 | |
| AnitalkerVenue=ACM MM 20242025.10 | 14.76 | 4.58 | 0.23 | 0.59 | 28.61 | 43.45 | 0.0053 | |
| Real3D-portraitVenue=ICLR 20242025.10 | 16.6 | 3.24 | 0.22 | 0.54 | 29.5 | 42.14 | 0.0062 |