Speech audio synthesis dataset 29 subjects (leave-one-out evaluation)
0.742Corr2DPLT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PLT2023.09 | 0.742 | 1.581 | |
| PLTAblation=w/o GAN2023.09 | 0.729 | 1.546 | |
| PLTAblation=w/o Pair-wise Disentangle2023.09 | 0.724 | 1.548 | |
| PLTModification=with cross embedding2023.09 | 0.72 | 1.55 | |
| LightViTProcessing=Padding 02023.09 | 0.708 | 1.475 | |
| LightViTProcessing=Bi-Cubic2023.09 | 0.702 | 1.492 | |
| CNN+SSPP2023.09 | 0.692 | 1.455 | |
| CNNProcessing=Bi-Cubic2023.09 | 0.689 | 1.451 | |
| CNNProcessing=Padding 02023.09 | 0.684 | 1.437 | |
| LightViTProcessing=Crop2023.09 | 0.635 | 1.208 | |
| CNNProcessing=Crop2023.09 | 0.614 | 1.126 |