Soundscape-to-image generation on SonicUrban 237K
11.553FIDSounDiT
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| SounDiT2025.05 | 11.553 | 0.52 | 0.706 | 0.52 | 0.739 | 0.759 | |
| PixArt + MHCA2025.05 | 41.456 | 0.517 | 0.592 | 0.411 | 0.396 | 0.796 | |
| Sound2Scene2025.05 | 49.681 | 0.503 | 0.498 | 0.332 | 0.324 | 0.829 | |
| GlueGen2025.05 | 64.434 | 0.498 | 0.499 | 0.405 | 0.481 | 0.805 | |
| CoDiPre-trained=true2025.05 | 83.904 | 0.503 | 0.497 | 0.273 | 0.31 | 0.856 | |
| AudioTokenBackbone/Base Model=SD12025.05 | 95.21 | 0.505 | 0.503 | 0.419 | 0.484 | 0.979 | |
| AudioTokenPre-trained=true2025.05 | 127.337 | 0.497 | 0.501 | 0.203 | 0.199 | 0.976 | |
| AudioTokenBackbone/Base Model=SD22025.05 | 165.014 | 0.502 | 0.504 | 0.298 | 0.455 | 0.98 |