MEAD
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Speech-Preserving Facial Expression Manipulation | MEAD (intra-identity) | FAD0.572 | 48 | |
| Speech-Preserving Facial Expression Manipulation | MEAD (cross-identity) | FAD1.872 | 38 | |
| Talking Head Generation | MEAD | Realism Score72 | 32 | |
| Audio-Driven Facial Animation | MEAD 41 (test) | PSNR31.873 | 26 | |
| Emotional Talking Head Generation | MEAD Cross-ID | Score (Neutral)45.2 | 24 | |
| Self-reenactment portrait animation | MEAD 59 (test) | CSIM0.9521 | 18 | |
| Audio-driven talking head generation | MEAD | Sync8.7778 | 14 | |
| Emotional Talking Head Generation | MEAD Intra-ID | Fidelity: Neutral9.452 | 12 | |
| Portrait Image Animation | MEAD (test) | FID32.696 | 12 | |
| 3D Face Reconstruction | MEAD | R Eye Error58.09 | 9 | |
| Audio-driven Video Generation | MEAD | FID3.77 | 8 | |
| Talking Head Generation | MEAD | FID45.403 | 8 | |
| Dynamic Facial Expression Recognition | MEAD 8-class | WAR88.44 | 8 | |
| Talking Head Generation | MEAD | FID41.81 | 7 | |
| 3D GAN Inversion | MEAD (novel views) | LPIPS (±60°)0.223 | 7 | |
| Audio-driven talking head synthesis | Mead 60 (test) | LSE-C1.76 | 7 | |
| Talking Head Generation | MEAD | Avg-R4.16 | 6 | |
| Talking Face Generation | MEAD | FID0 | 6 | |
| 3D Talking Face Synthesis | Emotional talking face dataset (MEAD) M003 and M030 (test) | Sync-C8.163 | 6 | |
| Emotional Talking Head Generation | MEAD | Emotion Score4.65 | 6 | |
| Facial Component Control | MEAD 53 (test) | Pose Error13.695 | 5 | |
| Emotion Editing | MEAD | AITV12.575 | 5 | |
| Video Compression | MEAD | BD-Rate (DISTS)16.6 | 4 | |
| Expression Control Accuracy | Mead | MSE0.188 | 4 | |
| Talking Face Generation | MEAD (test) | Mean Lip Deviation2.45 | 4 |
Showing 25 of 26 rows