3D Audio-Visual Scene Generation on SONOSCENE360
67.6D-CLAP Score (R Component)SONOWORLD
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SONOWORLDversion=Proprietary2026.03 | 67.6 | 0.672 | 0.216 | 0.728 | 0.658 | 0.838 | 48 | 45.7 | |
| SONOWORLDversion=Open-source2026.03 | 52.9 | 0.976 | 0.251 | 0.975 | 0.491 | 0.753 | 46.4 | 41.3 | |
| OmniAudio2026.03 | 39.7 | 1.508 | 0.317 | 1.449 | 0.148 | 0.588 | 10.1 | 10.4 | |
| MMAudio2026.03 | 33.8 | — | — | — | — | — | 34.5 | 32.2 | |
| SEE-2-SOUND2026.03 | 22.1 | 1.354 | 0.254 | 1.397 | 0.194 | 0.603 | 22.1 | 15.6 | |
| ViSAGe2026.03 | 22.1 | 1.598 | 0.426 | 1.649 | 0.142 | 0.624 | 8.8 | 1.9 |