Research

Paper

AI LLM March 11, 2026

Distilling LLM Semantic Priors into Encoder-Only Multi-Talker ASR with Talker-Count Routing

Authors

Hao Shi, Yusuke Fujita, Roman Koshkin, Mengjie Zhao, Yuan Gao, Lianbo Liu, Yui Sudo

Abstract

Large language models (LLMs) provide strong semantic priors that can improve multi-talker automatic speech recognition (MT-ASR), but using an LLM as an autoregressive decoder is computationally expensive and remains fragile under heavy overlap. In this paper, we propose an encoder-only MT-ASR framework that adapts an LLM to multi-talker conditioning and distills its semantic guidance into the encoder during training, while retaining fast CTC-style decoding at inference. Our model employs a post-encoder separator with serialized CTC to produce talker-ordered transcripts, and leverages an adapted LLM-based SOT objective as a multi-talker-aware teacher signal to explicitly regularize mixed-speech representations. To further support variable numbers of talkers, we introduce a Talker-Count Head that predicts the talker count and dynamically selects the appropriate decoding branch. Experiments on LibriMix show that the proposed encoder-only model achieves comparable performance to LLM-based systems in the two-talker condition, while delivering significant improvements in the three-talker condition with significant small RTF.

Metadata

arXiv ID: 2603.10587
Provider: ARXIV
Primary Category: cs.SD
Published: 2026-03-11
Fetched: 2026-03-12 04:21

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.10587v1</id>\n    <title>Distilling LLM Semantic Priors into Encoder-Only Multi-Talker ASR with Talker-Count Routing</title>\n    <updated>2026-03-11T09:43:49Z</updated>\n    <link href='https://arxiv.org/abs/2603.10587v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.10587v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Large language models (LLMs) provide strong semantic priors that can improve multi-talker automatic speech recognition (MT-ASR), but using an LLM as an autoregressive decoder is computationally expensive and remains fragile under heavy overlap. In this paper, we propose an encoder-only MT-ASR framework that adapts an LLM to multi-talker conditioning and distills its semantic guidance into the encoder during training, while retaining fast CTC-style decoding at inference. Our model employs a post-encoder separator with serialized CTC to produce talker-ordered transcripts, and leverages an adapted LLM-based SOT objective as a multi-talker-aware teacher signal to explicitly regularize mixed-speech representations. To further support variable numbers of talkers, we introduce a Talker-Count Head that predicts the talker count and dynamically selects the appropriate decoding branch. Experiments on LibriMix show that the proposed encoder-only model achieves comparable performance to LLM-based systems in the two-talker condition, while delivering significant improvements in the three-talker condition with significant small RTF.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.SD'/>\n    <published>2026-03-11T09:43:49Z</published>\n    <arxiv:primary_category term='cs.SD'/>\n    <author>\n      <name>Hao Shi</name>\n    </author>\n    <author>\n      <name>Yusuke Fujita</name>\n    </author>\n    <author>\n      <name>Roman Koshkin</name>\n    </author>\n    <author>\n      <name>Mengjie Zhao</name>\n    </author>\n    <author>\n      <name>Yuan Gao</name>\n    </author>\n    <author>\n      <name>Lianbo Liu</name>\n    </author>\n    <author>\n      <name>Yui Sudo</name>\n    </author>\n  </entry>"
}