A Comparative Study of Self-Supervised Speech Representations in Read and Spontaneous TTS
It addresses the selection of speech representations for TTS, particularly for challenging spontaneous speech, but is incremental as it compares existing methods.
This study compared self-supervised speech representations in text-to-speech (TTS) systems for read and spontaneous speech, finding that the 9th layer of wav2vec2.0 outperformed other representations and mel-spectrograms in listening tests.
Recent work has explored using self-supervised learning (SSL) speech representations such as wav2vec2.0 as the representation medium in standard two-stage TTS, in place of conventionally used mel-spectrograms. It is however unclear which speech SSL is the better fit for TTS, and whether or not the performance differs between read and spontaneous TTS, the later of which is arguably more challenging. This study aims at addressing these questions by testing several speech SSLs, including different layers of the same SSL, in two-stage TTS on both read and spontaneous corpora, while maintaining constant TTS model architecture and training settings. Results from listening tests show that the 9th layer of 12-layer wav2vec2.0 (ASR finetuned) outperforms other tested SSLs and mel-spectrogram, in both read and spontaneous TTS. Our work sheds light on both how speech SSL can readily improve current TTS systems, and how SSLs compare in the challenging generative task of TTS. Audio examples can be found at https://www.speech.kth.se/tts-demos/ssr_tts