Multimodal language models that process both text and speech have a potential for applications in spoken dialogue systems. However, current models face two major challenges in response generation latency: (1) generating a spoken response requires the prior generation of a written response, and (2) speech sequences are significantly longer than text sequences. This study addresses these issues by extending the input and output sequences of the language model to support the parallel generation of text and speech. Our experiments on spoken question answering tasks demonstrate that our approach improves latency while maintaining the quality of response content. Additionally, we show that latency can be further reduced by generating speech in multiple sequences. Demo samples are available at https://rinnakk.github.io/research/publications/PSLM.
@article{arxiv.2406.12428,
title = {PSLM: Parallel Generation of Text and Speech with LLMs for Low-Latency Spoken Dialogue Systems},
author = {Kentaro Mitsui and Koh Mitsuda and Toshiaki Wakatsuki and Yukiya Hono and Kei Sawada},
journal= {arXiv preprint arXiv:2406.12428},
year = {2024}
}
Comments
9 pages, 6 figures, 4 tables, accepted for Findings of EMNLP 2024. Demo samples: https://rinnakk.github.io/research/publications/PSLM