Zero-shot Text-to-Speech on MLS Es filtered (test)
3.5WERVoicebox (Multilingual)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Voicebox (Multilingual)Ref (audio context language)=Pl, alpha=12023.06 | 3.5 | 0.528 | |
| Voicebox (Multilingual)Ref (audio context language)=De, alpha=12023.06 | 3.6 | 0.442 | |
| Voicebox (Multilingual)Ref (audio context language)=Es, alpha=12023.06 | 3.6 | 0.613 | |
| Voicebox (Multilingual)Ref (audio context language)=Fr, alpha=12023.06 | 3.7 | 0.479 | |
| Voicebox (Multilingual)Ref (audio context language)=Pt, alpha=12023.06 | 3.7 | 0.476 | |
| Voicebox (Multilingual)Ref (audio context language)=AVG, alpha=12023.06 | 3.7 | 0.494 | |
| Ground Truth2023.06 | 4.1 | 0.729 | |
| Voicebox (Multilingual)Ref (audio context language)=En, alpha=12023.06 | 4.1 | 0.423 |