Transforming Spectrum and Prosody for Emotional Voice Conversion with Non-Parallel Training Data
This addresses the problem of generating emotional speech for applications like human-computer interaction, but it is incremental as it builds on existing non-parallel methods with a novel F0 modeling approach.
The paper tackled emotional voice conversion without parallel training data by proposing a CycleGAN framework and modeling fundamental frequency (F0) with continuous wavelet transform, achieving superior performance in objective and subjective evaluations compared to baselines.
Emotional voice conversion aims to convert the spectrum and prosody to change the emotional patterns of speech, while preserving the speaker identity and linguistic content. Many studies require parallel speech data between different emotional patterns, which is not practical in real life. Moreover, they often model the conversion of fundamental frequency (F0) with a simple linear transform. As F0 is a key aspect of intonation that is hierarchical in nature, we believe that it is more adequate to model F0 in different temporal scales by using wavelet transform. We propose a CycleGAN network to find an optimal pseudo pair from non-parallel training data by learning forward and inverse mappings simultaneously using adversarial and cycle-consistency losses. We also study the use of continuous wavelet transform (CWT) to decompose F0 into ten temporal scales, that describes speech prosody at different time resolution, for effective F0 conversion. Experimental results show that our proposed framework outperforms the baselines both in objective and subjective evaluations.