Speech-to-Speech Translation on CVSS-C Es-En v1 (test)
32.3ASR-BLEUUnitY
Evaluation Results
| Method | Links | |
|---|---|---|
| UnitYTYPE=Discrete2023.04 | 32.3 | |
| ESPnet-ST-v2 D (Multi-Decoder)TYPE=Discrete, Architecture=UnitY-like2023.04 | 32 | |
| ESPnet-ST-v2 C (Attn Enc-Dec)TYPE=Discrete, Architecture=Speech-to-Unit-like2023.04 | 31.3 | |
| Speech-to-UnitTYPE=Discrete2023.04 | 30.8 | |
| Translatotron2TYPE=Spectral2023.04 | 30.3 | |
| ESPnet-ST-v2 B (Multi-Decoder)TYPE=Spectral, Architecture=Translatotron2-like2023.04 | 24.3 | |
| ESPnet-ST-v2 A (Attn Enc-Dec)TYPE=Spectral, Architecture=Translatotron-like2023.04 | 16.6 | |
| TranslatotronTYPE=Spectral2023.04 | 14.4 |