List your company

Emotional Text-to-Speech on IEMOCAP (high-mismatch set)

0.874E-SIMCoCoEmo
0.799120.818560.8380.85744Feb 3, 2026

Evaluation results

MethodLinks
CoCoEmoBackbone=IndexTTS2, alpha=6.02026.020.8740.6810.8866.584.21
Emo vectorBackbone=IndexTTS22026.020.8720.6670.8775.754.05
CoCoEmoBackbone=CosyVoice2, alpha=6.02026.020.8620.5040.8928.744.4
CoCoEmoBackbone=IndexTTS2, alpha=3.02026.020.8570.4780.8845.344.08
InstructionBackbone=CosyVoice22026.020.8430.4360.8874.014.23
CoCoEmoBackbone=CosyVoice2, alpha=3.02026.020.8330.3320.8958.034.29
No-steerBackbone=IndexTTS22026.020.8250.3180.8855.134.17
No-steerBackbone=CosyVoice22026.020.8020.1970.8969.184.22