Emotional Text-To-Speech Based on Mutual-Information-Guided Emotion-Timbre Disentanglement
Abstract
Current emotional Text-To-Speech (TTS) and style transfer methods rely on reference encoders to control global style or emotion vectors, but do not capture nuanced acoustic details of the reference speech. To this end, we propose a novel emotional TTS method that enables fine-grained phoneme-level emotion embedding prediction while disentangling intrinsic attributes of the reference speech. The proposed method employs a style disentanglement method to guide two feature extractors, reducing mutual information between timbre and emotion features, and effectively separating distinct style components from the reference speech. Experimental results demonstrate that our method outperforms baseline TTS systems in generating natural and emotionally rich speech. This work highlights the potential of disentangled and fine-grained representations in advancing the quality and flexibility of emotional TTS systems.
Keywords
Cite
@article{arxiv.2510.01722,
title = {Emotional Text-To-Speech Based on Mutual-Information-Guided Emotion-Timbre Disentanglement},
author = {Jianing Yang and Sheng Li and Takahiro Shinozaki and Yuki Saito and Hiroshi Saruwatari},
journal= {arXiv preprint arXiv:2510.01722},
year = {2025}
}
Comments
In Proceedings of the 17th Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC 2025)