Image-to-Video Generation on MSR-VTT
91.3FVDTRIP
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TRIPModel Type=I2V, Zero-shot=true2024.03 | 91.3 | 9.68 | — | — | |
| VideoComposerModel Type=I2V, Zero-shot=true2024.03 | 208 | 31.29 | — | — | |
| SVD+CV-VAE-ITrainable=No, FCR=1x, Frames=252024.05 | 262 | — | 1.67 | 0.609 | |
| SVD+CV-VAE-VTrainable=Output layer, FCR=4x, Frames=972024.05 | 295 | — | 2.26 | 0.734 | |
| SVDTrainable=No, FCR=1x, Frames=252024.05 | 310 | — | 1.3 | 0.588 | |
| SVD+CV-VAE-VTrainable=No, FCR=4x, Frames=972024.05 | 319 | — | 3.31 | 0.696 |