DuplexChat: Constructing Speaker-Separated Full-Duplex Dialogue Speech at Scale for Spoken Dialogue Language Modeling
2026-07-06 • Computation and Language
Computation and LanguageSound
AI summaryⓘ
The authors created DuplexChat, a new dataset designed to help train dialogue systems that can listen and respond like humans in conversations. They made a process called DuplexChat-Pipe that takes public podcast audio and separates the voices of two speakers into different channels. This way, the data mimics real conversations with individual speaker streams, which most existing datasets lack. Their work resulted in a large English and Japanese dataset that includes natural turn-taking patterns found in real human dialogues.
full-duplex speechspoken dialogue modelsmonaural audiopodcast audiospeaker diarizationspeech separationspeech restorationturn-takingcorpus constructiondialogue systems
Authors
Wataru Nakata, Yuki Saito, Hiroshi Saruwatari
Abstract
Full-duplex spoken dialogue models are trained on conversational speech in which each speaker is represented as a separate stream, but existing large-scale public speech corpora are mostly monaural, making them unsuited for SDLM training. We present DuplexChat, an open-source corpus for full-duplex spoken dialogue models, and DuplexChat-Pipe, a pipeline for constructing speaker-separated full-duplex dialogue speech from public podcast feeds. DuplexChat-Pipe filters language-specific podcast feeds, retrieves and cleans episode audio, extracts diarization-guided two-speaker dialogue clips, and applies speech separation and restoration to produce one channel per speaker. Running this pipeline yields a speaker-separated spoken dialogue corpus covering 282,634 hours of English and 132,723 hours of Japanese. Analysis results on DuplexChat show that it contains turn-taking dynamics present in human dialogues.