Multimodal motion generation on Aff-Wild facial expression coefficients (train)
7.9S-MSEVanilla VAE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Vanilla VAE2018.08 | 7.9 | 3.2 | |
| MT-VAEfusion=concat2018.08 | 10.1 | 2.2 | |
| MT-VAEfusion=add2018.08 | 10.2 | 2 | |
| Prediction LSTM2018.08 | 15.3 | — | |
| Sequence Motion2018.08 | 184 | — | |
| Last-step Motion2018.08 | 638 | — |