Voice Cloning on LibriHeavy-HQ
40.778SNRLM-Backbone LoRA for TTS
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| LM-Backbone LoRA for TTSSpeaker ID=1401, Dataset Length (Hr)=17.5902026.03 | 40.778 | 54.6 | 3.749 | — | 52.207 | 26.787 | |
| NeuTTSSpeaker ID=1401, Dataset Length (Hr)=17.5902026.03 | 32.163 | 43.8 | 3.659 | 20.049 | — | — | |
| Reference AudioSpeaker ID=1401, Dataset Length (Hr)=17.5902026.03 | 26.791 | — | 3.564 | — | — | — | |
| Reference AudioSpeaker ID=1212, Dataset Length (Hr)=18.4922026.03 | 22.032 | — | 3.242 | — | — | — | |
| NeuTTSSpeaker ID=1212, Dataset Length (Hr)=18.4922026.03 | 20.397 | 51.1 | 3.647 | -7.42 | — | — | |
| Reference AudioSpeaker ID=1259, Dataset Length (Hr)=13.6522026.03 | 17.63 | — | 3.772 | — | — | — | |
| NeuTTSSpeaker ID=1259, Dataset Length (Hr)=13.6522026.03 | 16.895 | 45 | 3.665 | -4.17 | — | — | |
| LM-Backbone LoRA for TTSSpeaker ID=1259, Dataset Length (Hr)=13.6522026.03 | 15.395 | 45.5 | 3.859 | — | -12.679 | -8.879 | |
| LM-Backbone LoRA for TTSSpeaker ID=1212, Dataset Length (Hr)=18.4922026.03 | 12.412 | 61 | 3.713 | — | -43.665 | -39.15 |