Speech-to-Singing on PopBuTFy (test)
2.8974LSDGT Mel
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GT MelType=Ground Truth Mel-spectrogram2023.05 | 2.8974 | 0.9959 | 0.3693 | — | — | |
| AlignSTSNote=main proposed system2023.05 | 5.0129 | 0.9934 | 0.5366 | — | — | |
| AlignSTSOptimization=GAN2023.05 | 5.4926 | 0.9875 | 0.5709 | — | — | |
| AlignSTSInference Protocol=zero-shot2023.05 | 5.6607 | 0.9871 | 0.5693 | — | — | |
| SpeechSplit 2.0Variant=without Speaker Encoder2023.05 | 5.7681 | 0.987 | 0.8262 | — | — | |
| Parekh et al.2023.05 | 7.3613 | 0.9218 | 0.7865 | — | — |