Neural Voice Cloning with a Few Samples
This addresses the problem of personalized speech interfaces for users needing quick voice cloning, but it is incremental as it builds on existing multi-speaker generative models.
The paper tackles the problem of voice cloning with limited audio samples, introducing two neural approaches: speaker adaptation and speaker encoding. The result shows that both methods achieve good performance in naturalness and similarity, with speaker adaptation offering better quality but speaker encoding being more efficient for low-resource deployment.
Voice cloning is a highly desired feature for personalized speech interfaces. Neural network based speech synthesis has been shown to generate high quality speech for a large number of speakers. In this paper, we introduce a neural voice cloning system that takes a few audio samples as input. We study two approaches: speaker adaptation and speaker encoding. Speaker adaptation is based on fine-tuning a multi-speaker generative model with a few cloning samples. Speaker encoding is based on training a separate model to directly infer a new speaker embedding from cloning audios and to be used with a multi-speaker generative model. In terms of naturalness of the speech and its similarity to original speaker, both approaches can achieve good performance, even with very few cloning audios. While speaker adaptation can achieve better naturalness and similarity, the cloning time or required memory for the speaker encoding approach is significantly less, making it favorable for low-resource deployment.