Recent TTS models with decoder-only Transformer architecture, such as SPEAR-TTS and VALL-E, achieve impressive naturalness and demonstrate the ability for zero-shot adaptation given a speech prompt. However, such decoder-only TTS models lack monotonic alignment constraints, sometimes leading to hallucination issues such as mispronunciation, word skipping and repeating. To address this limitation, we propose VALL-T, a generative Transducer model that introduces shifting relative position embeddings for input phoneme sequence, explicitly indicating the monotonic generation process while maintaining the architecture of decoder-only Transformer. Consequently, VALL-T retains the capability of prompt-based zero-shot adaptation and demonstrates better robustness against hallucinations with a relative reduction of 28.3% in the word error rate.
@article{arxiv.2401.14321,
title = {VALL-T: Decoder-Only Generative Transducer for Robust and Decoding-Controllable Text-to-Speech},
author = {Chenpeng Du and Yiwei Guo and Hankun Wang and Yifan Yang and Zhikang Niu and Shuai Wang and Hui Zhang and Xie Chen and Kai Yu},
journal= {arXiv preprint arXiv:2401.14321},
year = {2025}
}