Talking Head Generation on HDTF
0FIDGround Truth
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ground Truth2026.05 | 0 | — | — | — | — | — | — | — | — | — | 1 | 7.78 | |
| MuseTalk2026.05 | 7.25 | — | — | — | — | — | — | — | — | — | 0.86 | 6.21 | |
| HighSync2026.05 | 7.36 | — | — | — | — | — | — | — | — | — | 0.85 | 7.72 | |
| SoulX-FlashHead (Pro)Streaming support=false2026.02 | 8.31 | 103.14 | 6.04 | 8.46 | — | — | — | — | — | — | — | — | |
| LatentSync2026.05 | 8.98 | — | — | — | — | — | — | — | — | — | 0.84 | 7.58 | |
| EchoMimicStreaming support=true2026.02 | 9 | 155.71 | 3.56 | 10.22 | 0.81 | — | — | — | — | — | — | — | |
| SoulX-FlashHead (Pro)Streaming support=true2026.02 | 9.97 | 111.38 | 5.73 | 8.77 | 10.81 | — | — | — | — | — | — | — | |
| SoulX-LiveAct2026.03 | 10.05 | 69.43 | 9.4 | 6.76 | — | — | 97.6 | 63 | 99.9 | — | — | — | |
| SoulX-FlashHead (Lite)Streaming support=false2026.02 | 10.78 | 115.94 | 5.12 | 8.8 | — | — | — | — | — | — | — | — | |
| SoulX-FlashHead (Lite)Streaming support=true2026.02 | 11.37 | 126.52 | 4.21 | 9.49 | 96 | — | — | — | — | — | — | — | |
| JAM-Flow (V2V)Inference (20s)=45 sec, #Params=∼500M (joint)2025.06 | 11.633 | 25.07 | 8.086 | 7.181 | — | — | — | — | — | — | — | — | |
| VideoReTalking2026.05 | 12.04 | — | — | — | — | — | — | — | — | — | 0.81 | 7.01 | |
| Wav2Lip2026.05 | 12.25 | — | — | — | — | — | — | — | — | — | 0.82 | 7.38 | |
| DittoStreaming support=true2026.02 | 12.35 | 199.13 | 3.57 | 10.49 | 45.04 | — | — | — | — | — | — | — | |
| Diff2Lip2026.05 | 13.45 | — | — | — | — | — | — | — | — | — | 0.79 | 7.08 | |
| SonicStreaming support=false2026.02 | 13.53 | 113.31 | 5.17 | 8.69 | — | — | — | — | — | — | — | — | |
| RESTStreaming=true, Runtime (s)=4.4162025.12 | 14.597 | 219.87 | 8.335 | 6.701 | — | 0.931 | — | — | — | — | — | — | |
| DittoStreaming=false, Runtime (s)=17.9742025.12 | 15.44 | 399.965 | 5.458 | 9.565 | — | 2.659 | — | — | — | — | — | — | |
| Live-Avatar2026.03 | 15.85 | 206.2 | 7.68 | 8.38 | — | — | 91.8 | 59.2 | 99.8 | — | — | — | |
| HalloStreaming=false, Runtime (s)=212.0022025.12 | 15.929 | 315.904 | 6.995 | 7.819 | — | 0.931 | — | — | — | — | — | — | |
| Hallo3Streaming support=true2026.02 | 15.95 | 160.94 | 3.18 | 10.72 | 0.16 | — | — | — | — | — | — | — | |
| DittoParams=0.1B2026.03 | 16.253 | 384.232 | 4.036 | 10.25 | — | 2.861 | — | — | — | 3 | — | — | |
| FantasyStreaming=false, Runtime (s)=896.0892025.12 | 16.489 | 315.291 | 5.138 | 10.349 | — | 1.232 | — | — | — | — | — | — | |
| SonicStreaming=false, Runtime (s)=83.5842025.12 | 16.894 | 245.416 | 8.525 | 6.576 | — | 0.932 | — | — | — | — | — | — | |
| JAM-Flow (I2V)Inference (20s)=45 sec, #Params=∼500M (joint)2025.06 | 17.571 | 192.3 | 7.324 | 7.777 | — | — | — | — | — | — | — | — | |
| AniPortraitStreaming=false, Runtime (s)=76.7782025.12 | 17.603 | 503.622 | 3.555 | 10.83 | — | 2.323 | — | — | — | — | — | — | |
| AniPortraitParams=2B2026.03 | 17.629 | 443.902 | 3.368 | 10.712 | — | 2.21 | — | — | — | 3.5 | — | — | |
| InfiniteTalk2026.03 | 18.15 | 169.88 | 7.12 | 8.01 | — | — | 94.5 | 59.9 | 99.4 | — | — | — | |
| JAM-Flow (Stage1)2025.06 | 18.372 | 194.27 | 7.138 | 7.947 | — | — | — | — | — | — | — | — | |
| EchoMimicStreaming=false, Runtime (s)=124.1052025.12 | 18.384 | 557.809 | 5.852 | 9.052 | — | 0.927 | — | — | — | — | — | — | |
| EchoMimicParams=2B2026.03 | 18.733 | 629.37 | 5.61 | 8.953 | — | 0.803 | — | — | — | 2.66 | — | — | |
| OursParams=0.6B2026.03 | 18.981 | 363.909 | 5.707 | 8.903 | — | 1.326 | — | — | — | 1.66 | — | — | |
| AniportraitStreaming support=false2026.02 | 19.83 | 242.29 | 1.89 | 11.91 | — | — | — | — | — | — | — | — | |
| Hallo32024.12 | 20.359 | 160.838 | 7.252 | 8.106 | — | — | — | — | — | — | — | — | |
| Hallo3Inference (20s)=30 min‡, #Params=1–3B2025.06 | 20.359 | 160.838 | 7.252 | 8.106 | — | — | — | — | — | — | — | — | |
| Hallo2024.12 | 20.545 | 173.497 | 7.75 | 7.659 | — | — | — | — | — | — | — | — | |
| HalloInference (20s)=23 min, #Params=1–3B2025.06 | 20.545 | 173.497 | 7.75 | 7.659 | — | — | — | — | — | — | — | — | |
| EchoMimicV3Params=1.3B2026.03 | 21.054 | 380.812 | 2.824 | 11.794 | — | 1.987 | — | — | — | 3.83 | — | — | |
| SadTalkerStreaming support=true2026.02 | 21.58 | 207.67 | 4.6 | 9.21 | 2.17 | — | — | — | — | — | — | — | |
| SadTalker2024.12 | 22.34 | 203.86 | 7.885 | 7.545 | — | — | — | — | — | — | — | — | |
| SadTalkerInference (20s)=2.5 min†, #Params=∼200M2025.06 | 22.34 | 203.86 | 7.885 | 7.545 | — | — | — | — | — | — | — | — | |
| AniPortrait2024.12 | 26.561 | 234.666 | 4.015 | 10.548 | — | — | — | — | — | — | — | — | |
| AniPortraitInference (20s)=7 min, #Params=∼1B2025.06 | 26.561 | 234.666 | 4.015 | 10.548 | — | — | — | — | — | — | — | — | |
| OmniAvatar2026.03 | 27.9 | 268.47 | 5.13 | 10.19 | — | — | 86.1 | 61.6 | 96.8 | — | — | — | |
| AniTalkerParams=0.1B2026.03 | 34.644 | 476.71 | 5.639 | 8.588 | — | 1.551 | — | — | — | 2.83 | — | — | |
| AniTalkerStreaming=true, Runtime (s)=13.5772025.12 | 39.155 | 514.388 | 5.838 | 8.736 | — | 1.523 | — | — | — | — | — | — | |
| DreamTalk2024.12 | 78.147 | 790.66 | 6.376 | 8.364 | — | — | — | — | — | — | — | — | |
| DreamTalk2025.06 | 78.147 | 890.66 | 6.376 | 8.364 | — | — | — | — | — | — | — | — | |
| Ground Truth2025.06 | — | — | 8.7 | 6.597 | — | — | — | — | — | — | — | — |