Towards Robust Neural Vocoding for Speech Generation: A Survey
This is an incremental study that addresses robustness issues in neural vocoding for speech generation applications.
The paper investigates the robustness of neural vocoders under data distribution mismatches, finding that speaker variety is more critical than language for universal vocoders, with WaveNet and WaveRNN better for text-to-speech and Parallel WaveGAN for voice conversion.
Recently, neural vocoders have been widely used in speech synthesis tasks, including text-to-speech and voice conversion. However, when encountering data distribution mismatch between training and inference, neural vocoders trained on real data often degrade in voice quality for unseen scenarios. In this paper, we train four common neural vocoders, including WaveNet, WaveRNN, FFTNet, Parallel WaveGAN alternately on five different datasets. To study the robustness of neural vocoders, we evaluate the models using acoustic features from seen/unseen speakers, seen/unseen languages, a text-to-speech model, and a voice conversion model. We found out that the speaker variety is much more important for achieving a universal vocoder than the language. Through our experiments, we show that WaveNet and WaveRNN are more suitable for text-to-speech models, while Parallel WaveGAN is more suitable for voice conversion applications. Great amount of subjective MOS results in naturalness for all vocoders are presented for future studies.