Text-to-Video Generation on User Study (Wan2.1-T2V-1.3B Evaluation)
4.11Content ConsistencyFreeLOC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FreeLOCBase Model=Wan2.1-T2V-1.3B2026.03 | 4.11 | 3.95 | 3.78 | |
| FreelongBase Model=Wan2.1-T2V-1.3B2026.03 | 3.59 | 2.11 | 2.98 | |
| RIFLEXBase Model=Wan2.1-T2V-1.3B2026.03 | 3.31 | 1.09 | 1.89 | |
| Direct SamplingBase Model=Wan2.1-T2V-1.3B2026.03 | 3.27 | 1.02 | 1.92 | |
| FreeNoiseBase Model=Wan2.1-T2V-1.3B2026.03 | 1.2 | 2.88 | 2.68 | |
| Sliding WindowBase Model=Wan2.1-T2V-1.3B2026.03 | 1.17 | 3.25 | 2.34 |