Robust Speaker Cloning
Large artifacts for the companion GitHub repository. V2 trains a 6-layer robust speaker Transformer and telephone BWE front end while keeping CosyVoice2-0.5B frozen. On 21 paired real CosyVoice generations, ECAPA...
Model source
- Kind
- Other
- File size
- 102 MB
- Version
- vRevision 9adc957
- Updated
- Oct 19, 2026
- Downloads
- 20
- Source
- Hugging Face
Source description
Robust Speaker Cloning artifacts
Large artifacts for the companion GitHub repository. V2 trains a 6-layer robust speaker Transformer and telephone BWE front end while keeping CosyVoice2-0.5B frozen. On 21 paired real CosyVoice generations, ECAPA x-vector similarity was 0.6187 for V2 versus 0.5998 baseline (+0.0189 average). This small evaluation also contains negative HVAC and reverb deltas, so it is not a universal-improvement claim.
CosyVoice, CampPlus, ECAPA and AISHELL retain their upstream licenses.
Sources
1 source
Sources
1 sourceVerified Oct 19
Model artifacts
2 artifacts
Model artifacts
2 artifactsSource excerpts
3 excerpts
Source excerpts
3 excerpts--- language: - zh tags: - text-to-speech - voice-cloning - cosyvoice - speaker-verification license: mit --- # Robust Speaker Cloning artifacts Large artifacts for the companion GitHub repository. V2 trains a 6-layer robust speaker Transformer and telephone BWE front end while keeping CosyVoice2-0.5B frozen. On 21 paired real CosyVoice generations, ECAPA x-vector similarity was 0.6187 for V2 versus 0.5998 baseline (+0.0189 average). This small evaluation also contains negative HVAC and reverb deltas, so it is not a universal-improvement claim. CosyVoice, CampPlus, ECAPA and AISHELL retain their upstream licenses.
Source context: 20 downloads · 0 likes · Pipeline text-to-speech · Library cosyvoice · Repo jatshi/robust-speaker-cloning
Source context: 106 downloads · 0 likes · Pipeline text-to-speech · Library cosyvoice · Repo jatshi/robust-speaker-cloning
