Audio-driven Generation on TalkBench Short (10 s) 1.0 (test)
3.51ASESoulX-FlashTalk
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SoulX-FlashTalkReal-time inference support=true2025.12 | 3.51 | 4.79 | 1.47 | 11.56 | 99.22 | 98.1 | 99.61 | 99.52 | 32 | |
| Echomimic-V32025.12 | 3.45 | 4.7 | 0.89 | 12.81 | 98.75 | 95.98 | 99.54 | 99.44 | 0.53 | |
| DittoReal-time inference support=true2025.12 | 3.1 | 4.37 | 1.04 | 12.58 | 99.8 | 99.23 | 99.75 | 99.86 | 21.8 | |
| LiveAvatarReal-time inference support=true2025.12 | 3.1 | 3.25 | 1.01 | 12.1 | 98.27 | 97.48 | 99.25 | 98.86 | 20.88 | |
| InfinitetalkLoRA version=true2025.12 | 3.09 | 3.04 | 1.25 | 11.89 | 98.71 | 97.04 | 99.54 | 99.34 | 5.1 | |
| OminiAvatar2025.12 | 3.06 | 3.01 | 1.32 | 11.85 | 98.64 | 96.95 | 99.6 | 99.66 | 0.16 | |
| StableAvatar2025.12 | 3.05 | 3.01 | 0.78 | 12.12 | 98.34 | 96.46 | 99.44 | 99.06 | 0.64 |