End-to-end Emotional Speech Synthesis Using Style Tokens And Semi-supervised Training

Abstract

This paper proposes an end-to-end emotional speech synthesis (ESS) method which adopts global style tokens (GSTs) for semi-supervised training. This model is built based on the GST-Tacotron framework. The style tokens are defined to present emotion categories. A cross entropy loss function between token weights and emotion labels is designed to obtain the interpretability of style tokens utilizing the small portion of training data with emotion labels. Emotion recognition experiments confirm that this method can achieve one-to-one correspondence between style tokens and emotion categories effectively. Objective and subjective evaluation results show that our model outperforms the conventional Tacotron model for ESS when only 5% of training data has emotion labels. Its subjective performance is close to the Tacotron model trained using all emotion labels.

Stats

• citations51
• S2 citations—
• github stars0
• HF likes0
• heat score12.87
• arxiv keywu2019end

Related papers

• Style Tokens: Unsupervised Style Modeling, Control And Transfer In End-to-end Speech Synthesis (2018)0.00
• Predicting Expressive Speaking Style From Text In End-to-end Speech Synthesis (2018)14.11
• Exploring Speech Style Spaces With Language Models: Emotional TTS Without Emotion Labels (2024)0.00
• Self-supervised Context-aware Style Representation For Expressive Speech Synthesis (2022)6.34
• Is Style All You Need? Dependencies Between Emotion And Gst-based Speaker Recognition (2022)0.00
• Uncovering Latent Style Factors For Expressive Speech Synthesis (2017)0.00
• Boosting Multi-speaker Expressive Speech Synthesis With Semi-supervised Contrastive Learning (2023)5.24
• Controllable Emotion Transfer For End-to-end Speech Synthesis (2020)13.05