FastSpeech: Fast, Robust and Controllable Text to Speech
Yi Ren, Yangjun Ruan, Xu Tan, Tao Qin, Sheng Zhao, Zhou Zhao, Tie-Yan Liu

Project

• FastSpeech is the first fully parallel end-to-end speech synthesis model.
• Academic Impact: This work is included by many famous speech synthesis open-source projects, such as ESPNet . Our work are promoted by more than 20 media and forums, such as 机器之心、InfoQ.
• Industry Impact: FastSpeech has been deployed in Microsoft Azure TTS service and supports 49 more languages with state-of-the-art AI quality. It was also shown as a text-to-speech system acceleration example in NVIDIA GTC2020.