Long-form Video Soundtrack Generation on LVS Benchmark
0.224ImageBind AvgJenBridge
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| JenBridge2026.06 | 0.224 | 8.12 | 7.83 | 8.21 | 4.4 | 4.3 | 4.2 | 4.3 | |
| VidMuseProcessing Mode=concatenated short-clip2026.06 | 0.162 | 6.81 | 4.42 | 6.75 | 3.8 | 3.8 | 1.8 | 3.1 | |
| VidMuseProcessing Mode=entire long video2026.06 | 0.148 | 6.89 | 5.56 | 6.82 | 3.9 | 3.8 | 2.5 | 3.4 | |
| CMTProcessing Mode=concatenated short-clip2026.06 | 0.143 | 5.52 | 4.21 | 5.38 | 3.3 | 3.2 | 1.6 | 2.7 | |
| AudioX2026.06 | 0.132 | 6.55 | 4.35 | 6.41 | 3.8 | 3.7 | 1.7 | 3.1 | |
| LORIS2026.06 | 0.121 | 4.81 | 4.15 | 4.52 | 2.9 | 2.7 | 1.6 | 2.4 | |
| CMTProcessing Mode=entire long video2026.06 | 0.115 | 5.61 | 4.88 | 5.45 | 3.4 | 3 | 2 | 2.8 |