Evidential-TTS: High Fidelity Zero-Shot Text-to-Speech Using Evidential Deep Learning
Authors:
- 6 April 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
We propose Evidential-TTS, a novel zero-shot text-to-speech (TTS) system based on evidential deep learning (EDL). The model includes a length regulator to ensure precise alignment between phonemes and acoustic tokens. This module allows the evidential token generator to convert the aligned phoneme sequence into acoustic tokens using iterative parallel decoding (IPD). However, IPD often suffers from overconfidence when using categorical probabilities as confidence scores. To address this, we introduce model uncertainty into the sampling process, quantified through EDL optimization. This uncertainty provides a more reliable sampling path for high-quality speech generation. Experimental results show that Evidential-TTS outperforms existing models in terms of speech naturalness and intelligibility. An ablation study further demonstrates the importance of uncertainty estimation in guiding the sampling trajectory of IPD.Keywords
This publication has 10 references indexed in Scilit:
- High Fidelity Text-to-Speech Via Discrete Tokens Using Token Transducer and Group Masked Language Model2024
- Efficient Parallel Audio Generation Using Group Masked Language ModelingIEEE Signal Processing Letters, 2024
- Evidential Uncertainty Quantification: A Variance-Based Perspective2024
- Attention Is All You NeedarXiv, 2023
- Improved Masked Image Generation with Token-Critic2022
- WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech ProcessingIEEE Journal of Selected Topics in Signal Processing, 2022
- MaskGIT: Masked Generative Image Transformer2022
- SoundStream: An End-to-End Neural Audio CodecIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2021
- ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification2020
- LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speech2019