Cross-lingual Emotion TTS


Authors: Cheng Gong, Chunyu Qiang, Tianrui Wang, Yu Jiang, Yuheng Lu,
Ruihaojing, Xiaoxiao Miao, Xiaolei Zhang, Longbiao Wang, Jianwu Dang

College of Intelligence and Computing, Tianjin University, China

Institute of Artificial Intelligence (TeleAI), China Telecom

Duke Kunshan University, China

Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Guangdong

English speech result:

Input Text: 是的真是名副其实。

Method (Intral-Lingual) Target CN Speaker Biaobei (Cross-Lingual) Target EN Speaker LJSpeech
Neutral Angry Happy Sad Surprise Neutral Angry Happy Sad Surprise
M3
DiCLET-TTS
EMM-TTS
Reference speech

Chinese speech result:

Input Text: What do you think of this question?

Method (Cross-Lingual) Target CN Speaker (Intral-Lingual) Target EN Speaker
Neutral Angry Happy Sad Surprise Neutral Angry Happy Sad Surprise
M3
DiCLET-TTS
EMM-TTS
Reference speech