Table 13 Subjective evaluation results with standard deviation (std dev) of voice conversion in cross-lingual scenario. The results are listd as MOS/std dev
From: U2-VC: one-shot voice conversion using two-level nested U-structure
| Â | MOS (similarity)/std dev | MOS (naturalness)/std dev | ||||||
|---|---|---|---|---|---|---|---|---|
| Â | VCTK2VCC | VCC2VCTK | VCC2VCC | Average | VCTK2VCC | VCC2VCTK | VCC2VCC | Average |
AdaIN-VC | 2.04/0.38 | 1.81/0.60 | 1.90/0.38 | 1.92 | 2.24/0.25 | 1.91/0.33 | 2.00/0.34 | 2.05 |
AGAIN-VC | 2.94/0.45 | 2.47/0.58 | 2.68/0.44 | 2.70 | 3.19/0.50 | 2.74/0.48 | 2.58/0.60 | 2.84 |
U2-VC | 3.33/0.37 | 3.10/0.48 | 3.14/0.49 | 3.19 | 3.78/0.48 | 3.44/0.46 | 3.26/0.40 | 3.49 |