SLIDE: Integrating Speech Language Model with LLM for Spontaneous Spoken Dialogue Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lu, Haitian, Cheng, Gaofeng, Luo, Liuping, Zhang, Leying, Qian, Yanmin, Zhang, Pengyuan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916547277094912
author Lu, Haitian
Cheng, Gaofeng
Luo, Liuping
Zhang, Leying
Qian, Yanmin
Zhang, Pengyuan
author_facet Lu, Haitian
Cheng, Gaofeng
Luo, Liuping
Zhang, Leying
Qian, Yanmin
Zhang, Pengyuan
contents Recently, ``textless" speech language models (SLMs) based on speech units have made huge progress in generating naturalistic speech, including non-verbal vocalizations. However, the generated speech samples often lack semantic coherence. In this paper, we propose SLM and LLM Integration for spontaneous spoken Dialogue gEneration (SLIDE). Specifically, we first utilize an LLM to generate the textual content of spoken dialogue. Next, we convert the textual dialogues into phoneme sequences and use a two-tower transformer-based duration predictor to predict the duration of each phoneme. Finally, an SLM conditioned on the spoken phoneme sequences is used to vocalize the textual dialogue. Experimental results on the Fisher dataset demonstrate that our system can generate naturalistic spoken dialogue while maintaining high semantic coherence.
format Preprint
id arxiv_https___arxiv_org_abs_2501_00805
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SLIDE: Integrating Speech Language Model with LLM for Spontaneous Spoken Dialogue Generation
Lu, Haitian
Cheng, Gaofeng
Luo, Liuping
Zhang, Leying
Qian, Yanmin
Zhang, Pengyuan
Audio and Speech Processing
Computation and Language
Sound
Recently, ``textless" speech language models (SLMs) based on speech units have made huge progress in generating naturalistic speech, including non-verbal vocalizations. However, the generated speech samples often lack semantic coherence. In this paper, we propose SLM and LLM Integration for spontaneous spoken Dialogue gEneration (SLIDE). Specifically, we first utilize an LLM to generate the textual content of spoken dialogue. Next, we convert the textual dialogues into phoneme sequences and use a two-tower transformer-based duration predictor to predict the duration of each phoneme. Finally, an SLM conditioned on the spoken phoneme sequences is used to vocalize the textual dialogue. Experimental results on the Fisher dataset demonstrate that our system can generate naturalistic spoken dialogue while maintaining high semantic coherence.
title SLIDE: Integrating Speech Language Model with LLM for Spontaneous Spoken Dialogue Generation
topic Audio and Speech Processing
Computation and Language
Sound
url https://arxiv.org/abs/2501.00805