Speech-Driven Facial Animation on VOCASET (test)
6.288Evl ScoreVOCA
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| VOCA2026.04 | 6.288 | — | — | — | — | 2.237 | 2.471 | |
| Faceformer2026.04 | 5.506 | — | — | — | — | 2.153 | 2.352 | |
| CodeTalker2026.04 | 5.278 | — | — | — | — | 2.175 | 2.238 | |
| SelfTalk2026.04 | 5.128 | — | — | — | — | 2.155 | 1.973 | |
| PATS2026.04 | 5.079 | — | — | — | — | 2.122 | 1.964 | |
| MMTalker2026.04 | 4.971 | — | — | — | — | 2.103 | 2.025 | |
| FaceDiffuserA/B preference test baseline=VOCA2023.09 | — | 76.83 | 78.05 | — | 78.05 | — | — | |
| FaceDiffuserA/B preference test baseline=FaceFormer2023.09 | — | 49.38 | 46.91 | — | 51.85 | — | — | |
| FaceDiffuserA/B preference test baseline=CodeTalker2023.09 | — | 27.16 | 26.4 | — | 27.16 | — | — | |
| FaceDiffuserA/B preference test baseline=GT2023.09 | — | 23.81 | 33.33 | — | 27.38 | — | — | |
| MeshTalk2021.04 | — | — | — | 3.184 | — | — | — | |
| VOCAAudio encoder=DeepSpeech [17]2021.04 | — | — | — | 3.72 | — | — | — | |
| VOCAAudio encoder=Mel spectrograms + current paper audio encoder2021.04 | — | — | — | 3.472 | — | — | — |