Msdtron: a high-capability multi-speaker speech synthesis system for diverse data using characteristic information
This work addresses the problem of handling diverse speaker characteristics in speech synthesis for applications requiring high-quality multi-speaker output, representing an incremental improvement.
The paper tackles the challenge of modeling diverse multi-speaker speech data by proposing Msdtron, a system that uses an excitation spectrogram and conditional gated LSTM to reduce mel-spectrogram reconstruction error and improve subjective evaluation scores in speaker adaptation.
In multi-speaker speech synthesis, data from a number of speakers usually tend to have great diversity due to the fact that the speakers may differ largely in ages, speaking styles, emotions, and so on. It is important but challenging to improve the modeling capabilities for multi-speaker speech synthesis. To address the issue, this paper proposes a high-capability speech synthesis system, called Msdtron, in which 1) a representation of the harmonic structure of speech, called excitation spectrogram, is designed to directly guide the learning of harmonics in mel-spectrogram. 2) conditional gated LSTM (CGLSTM) is proposed to control the flow of text content information through the network by re-weighting the gates of LSTM using speaker information. The experiments show a significant reduction in reconstruction error of mel-spectrogram in the training of the multi-speaker model, and a great improvement is observed in the subjective evaluation of speaker adapted model.