Video Synthesis on iPER (test)
132.9FVDOurs
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Oursarchitecture=full model, components=cINN + x0 + AdaIN2021.05 | 132.9 | 0.5 | 1.63 | |
| Hierarchical Latent Model2021.06 | 144.92 | — | — | |
| SAVP2021.06 | 150.39 | — | — | |
| Ours w/o ADAINablation=exclude ADAIN2021.05 | 156.7 | 0.48 | 1.6 | |
| Ours w/o cINNablation=exclude cINN2021.05 | 180.6 | 0.32 | 1.21 | |
| IVRNN2021.06 | 181.5 | — | — | |
| IVRNNmode=stochastic video synthesis2021.05 | 191.4 | 0.23 | 0.57 | |
| SRVP2021.06 | 245.13 | — | — | |
| SRVPmode=stochastic video synthesis2021.05 | 336.3 | 0.34 | 1.01 | |
| SAVPmode=stochastic video synthesis2021.05 | 368.6 | 0 | 0.01 | |
| Ours w/o x0ablation=exclude x02021.05 | 381.5 | 0.73 | 2.15 |