Soundscape-to-image generation on SoundingSVI 169K
16.839FIDSounDiT
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| SounDiT2025.05 | 16.839 | 53.8 | 75.3 | 57.2 | 75.3 | 72.9 | |
| PixArt + MHCA2025.05 | 34.108 | 51.8 | 57.8 | 47.4 | 39 | 74.3 | |
| GlueGen2025.05 | 74.25 | 50 | 56.6 | 40 | 46.1 | 77.6 | |
| Sound2Scene2025.05 | 78.137 | 58.6 | 62.3 | 46.7 | 37.4 | 73.5 | |
| CoDiPre-trained=true2025.05 | 87.027 | 64.1 | 61.2 | 40 | 30.2 | 80 | |
| AudioTokenBackbone/Base Model=SD12025.05 | 105.472 | 50.9 | 57.2 | 52.7 | 55.3 | 79.7 | |
| AudioTokenPre-trained=true2025.05 | 142.441 | 60 | 51.9 | 24.8 | 18.4 | 88.4 | |
| AudioTokenBackbone/Base Model=SD22025.05 | 149.473 | 50.4 | 55.9 | 24.9 | 31.1 | 84.5 |