Timezone: »

Deep Voice: Real-time Neural Text-to-Speech
Andrew Gibiansky · Mike Chrzanowski · Mohammad Shoeybi · Shubho Sengupta · Gregory Diamos · Sercan Arik · Jonathan Raiman · John Miller · Xian Li · Yongguo Kang · Adam Coates · Andrew Ng

Tue Aug 08 01:30 AM -- 05:00 AM (PDT) @ Gallery #84

We present Deep Voice, a production-quality text-to-speech system constructed entirely from deep neural networks. Deep Voice lays the groundwork for truly end-to-end neural speech synthesis. The system comprises five major building blocks: a segmentation model for locating phoneme boundaries, a grapheme-to-phoneme conversion model, a phoneme duration prediction model, a fundamental frequency prediction model, and an audio synthesis model. For the segmentation model, we propose a novel way of performing phoneme boundary detection with deep neural networks using connectionist temporal classification (CTC) loss. For the audio synthesis model, we implement a variant of WaveNet that requires fewer parameters and trains faster than the original. By using a neural network for each component, our system is simpler and more flexible than traditional text-to-speech systems, where each component requires laborious feature engineering and extensive domain expertise. Finally, we show that inference with our system can be performed faster than real time and describe optimized WaveNet inference kernels on both CPU and GPU that achieve up to 400x speedups over existing implementations.

Author Information

agibiansky Gibiansky (Baidu Research Silicon Valley AI Lab)
Mike Chrzanowski (Baidu Research)
Mohammad Shoeybi (Baidu Research)
Shubho Sengupta (Baidu Research)
Gregory Diamos (Baidu Research)
Sercan Arik (Baidu Research)
Jonathan Raiman (Baidu Research)
John Miller (Baidu Research)
Xian Li (Baidu)
Yongguo Kang (Baidu)
Adam Coates (Baidu SVAIL)
Andrew Ng (Baidu)

Related Events (a corresponding poster, oral, or spotlight)