Emotional Speech Synthesis on 48-Category Zero-Shot Dataset
87ECSEmoInstruct-TTS (Dual-Path)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| EmoInstruct-TTS (Dual-Path)zero-shot=true2026.06 | 87 | 2.59 | |
| EmoInstruct-TTS -w/o Text Instruct (Emo2emb-Only)zero-shot=true2026.06 | 86.7 | 4.86 | |
| CosyVoice3zero-shot=true2026.06 | 86.5 | 1.97 | |
| EmoInstruct-TTS -w/o Emo2emb (Text-Only)zero-shot=true2026.06 | 85.9 | 3.29 | |
| CosyVoice2zero-shot=true2026.06 | 85.5 | 3.57 |