Phrase break prediction with bidirectional encoder representations in Japanese text-to-speech synthesis
This work addresses improving naturalness in TTS systems for Japanese language applications, representing an incremental advancement.
The paper tackles phrase break prediction in Japanese text-to-speech synthesis by combining BERT and BiLSTM features, achieving a 3.2-point F1 score improvement over conventional methods and a mean opinion score of 4.39 for prosody naturalness, nearly matching ground-truth performance.
We propose a novel phrase break prediction method that combines implicit features extracted from a pre-trained large language model, a.k.a BERT, and explicit features extracted from BiLSTM with linguistic features. In conventional BiLSTM based methods, word representations and/or sentence representations are used as independent components. The proposed method takes account of both representations to extract the latent semantics, which cannot be captured by previous methods. The objective evaluation results show that the proposed method obtains an absolute improvement of 3.2 points for the F1 score compared with BiLSTM-based conventional methods using linguistic features. Moreover, the perceptual listening test results verify that a TTS system that applied our proposed method achieved a mean opinion score of 4.39 in prosody naturalness, which is highly competitive with the score of 4.37 for synthesized speech with ground-truth phrase breaks.