Spatial Audio Generation on YT360 (test)
145.83FDSwanSphere
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SwanSphereParams=1.09B, Inf. Time=0.21s/9.13s2026.05 | 145.83 | 1.6 | 1.24 | 0.45 | 1.13 | |
| OmniAudioParams=1.22B, Inf. Time=0.85s2026.05 | 186.42 | 2.17 | 1.42 | 0.55 | 1.44 | |
| ViSAGeParams=0.36B, Inf. Time=20.19s2026.05 | 266.91 | 2.97 | 1.82 | 0.75 | 1.86 | |
| MMAudio+ASParams=1.03B, Inf. Time=2.76s2026.05 | 276.84 | 2.64 | 1.42 | 0.57 | — | |
| Diff-Foley+ASParams=0.94B, Inf. Time=2.03s2026.05 | 314.65 | 2.95 | — | — | — |