Adapting Diarization-Conditioned Whisper for End-to-End Multi-Talker Speech Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kocour, Martin, Karafiat, Martin, Polok, Alexander, Klement, Dominik, Burget, Lukáš, Černocký, Jan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912874249584640
author Kocour, Martin
Karafiat, Martin
Polok, Alexander
Klement, Dominik
Burget, Lukáš
Černocký, Jan
author_facet Kocour, Martin
Karafiat, Martin
Polok, Alexander
Klement, Dominik
Burget, Lukáš
Černocký, Jan
contents We propose a speaker-attributed (SA) Whisper-based model for multi-talker speech recognition that combines target-speaker modeling with serialized output training (SOT). Our approach leverages a Diarization-Conditioned Whisper (DiCoW) encoder to extract target-speaker embeddings, which are concatenated into a single representation and passed to a shared decoder. This enables the model to transcribe overlapping speech as a serialized output stream with speaker tags and timestamps. In contrast to target-speaker ASR systems such as DiCoW, which decode each speaker separately, our approach performs joint decoding, allowing the decoder to condition on the context of all speakers simultaneously. Experiments show that the model outperforms existing SOT-based approaches and surpasses DiCoW on multi-talker mixtures (e.g., LibriMix).
format Preprint
id arxiv_https___arxiv_org_abs_2510_03723
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Adapting Diarization-Conditioned Whisper for End-to-End Multi-Talker Speech Recognition
Kocour, Martin
Karafiat, Martin
Polok, Alexander
Klement, Dominik
Burget, Lukáš
Černocký, Jan
Audio and Speech Processing
Computation and Language
Sound
We propose a speaker-attributed (SA) Whisper-based model for multi-talker speech recognition that combines target-speaker modeling with serialized output training (SOT). Our approach leverages a Diarization-Conditioned Whisper (DiCoW) encoder to extract target-speaker embeddings, which are concatenated into a single representation and passed to a shared decoder. This enables the model to transcribe overlapping speech as a serialized output stream with speaker tags and timestamps. In contrast to target-speaker ASR systems such as DiCoW, which decode each speaker separately, our approach performs joint decoding, allowing the decoder to condition on the context of all speakers simultaneously. Experiments show that the model outperforms existing SOT-based approaches and surpasses DiCoW on multi-talker mixtures (e.g., LibriMix).
title Adapting Diarization-Conditioned Whisper for End-to-End Multi-Talker Speech Recognition
topic Audio and Speech Processing
Computation and Language
Sound
url https://arxiv.org/abs/2510.03723