Multi-shot Audio-Visual Generation on MAVINSet subjective 20 samples
36.8AVQMAVIN
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| MAVINGeneration Strategy=Joint Generation Models: T2AV2026.06 | 36.8 | 42.2 | 66.4 | 64.6 | |
| CineTransGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 9.4 | 6.6 | 5 | 5.2 | |
| EchoShotGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 8.2 | 7 | 5.2 | 5.4 | |
| OVIGeneration Strategy=Joint Generation Models: T2AV2026.06 | 7.8 | 7.6 | 4.6 | 4.4 | |
| IC-LoRA + WanGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 6.6 | 4.8 | 2.8 | 3.2 | |
| LTX-2Generation Strategy=Joint Generation Models: T2AV2026.06 | 6.6 | 7.2 | 2.4 | 2.4 | |
| StoryDiff. + WanGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 6.2 | 5 | 3 | 3.4 | |
| VideoGen-of-ThoughtGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 5.6 | 8 | 3.2 | 3.4 | |
| MovieAgentGeneration Strategy=Cascaded Generation Models: Multi-shot T2V + V2A2026.06 | 4.8 | 4.6 | 2.6 | 2.8 | |
| UniVerse-1Generation Strategy=Joint Generation Models: T2AV2026.06 | 4.6 | 4 | 2.8 | 3 | |
| JavisDiTGeneration Strategy=Joint Generation Models: T2AV2026.06 | 3.4 | 3 | 2 | 2.2 |