Retrieval-Augmented Self-Taught Reasoning Model with Adaptive Chain-of-Thought for ASR Named Entity Correction

• URL: http://arxiv.org/abs/2602.12287v1
• Date: Wed, 21 Jan 2026 15:05:39 GMT
• Title: Retrieval-Augmented Self-Taught Reasoning Model with Adaptive Chain-of-Thought for ASR Named Entity Correction
• Authors: Junjie An, Jingguang Tian, Tianyi Wang, Yu Gao, Xiaofeng Mou, Yi Xu,
• Abstract summary: We propose a retrieval-augmented generation framework for correcting named entity errors in automatic speech recognition (ASR)<n>Our approach consists of two key components: (1) a rephrasing language model (RLM) for named entity recognition, followed by candidate retrieval using a phonetic-level edit distance; and (2) a novel self-taught reasoning model with adaptive chain-of-thought (A-STAR) that dynamically adjusts the depth of its reasoning based on task difficulty.
• Score: 12.483998165719981
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: End-to-end automatic speech recognition (ASR) systems frequently misrecognize domain-specific phrases like named entities, which can cause catastrophic failures in downstream tasks. A new family of named entity correction methods based on large language models (LLMs) has recently emerged. However, these approaches have yet to fully exploit the sophisticated reasoning capabilities inherent to LLMs. To bridge this gap, we propose a novel retrieval-augmented generation framework for correcting named entity errors in ASR. Our approach consists of two key components: (1) a rephrasing language model (RLM) for named entity recognition, followed by candidate retrieval using a phonetic-level edit distance; and (2) a novel self-taught reasoning model with adaptive chain-of-thought (A-STAR) that dynamically adjusts the depth of its reasoning based on task difficulty. Experiments on the AISHELL-1 and Homophone datasets demonstrate the effectiveness of our method, which achieves relative reductions in the named entity character error rate of 17.96\% and 34.42\%, respectively, compared to a strong baseline.

Related papers

• SUTA-LM: Bridging Test-Time Adaptation and Language Model Rescoring for Robust ASR [58.31068047426522]
  Test-Time Adaptation (TTA) aims to mitigate by adjusting models during inference.<n>Recent work explores combining TTA with external language models, using techniques like beam search rescoring or generative error correction.<n>We propose SUTA-LM, a simple yet effective extension of SUTA, with language model rescoring.<n> Experiments on 18 diverse ASR datasets show that SUTA-LM achieves robust results across a wide range of domains.
  arXiv  Detail & Related papers  (2025-06-10T02:50:20Z)
• Customizing Speech Recognition Model with Large Language Model Feedback [5.290365603660415]
  We propose a reinforcement learning based approach for unsupervised domain adaptation.<n>We leverage unlabeled data to enhance transcription quality, particularly the named entities affected by domain mismatch.<n>Our method achieves a 21% improvement on entity word error rate over conventional self-training methods.
  arXiv  Detail & Related papers  (2025-06-05T18:42:57Z)
• LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context [4.444835399672951]
  We propose a novel GER approach that targets rare words and incorporates phonetic information.<n> Experimental results show that our method not only improves the correction of rare words but also reduces the WER and CER.
  arXiv  Detail & Related papers  (2025-05-23T02:54:52Z)
• Failing Forward: Improving Generative Error Correction for ASR with Synthetic Data and Retrieval Augmentation [73.9145653659403]
  We show that Generative Error Correction models struggle to generalize beyond the specific types of errors encountered during training. We propose DARAG, a novel approach designed to improve GEC for ASR in in-domain (ID) and OOD scenarios. Our approach is simple, scalable, and both domain- and language-agnostic.
  arXiv  Detail & Related papers  (2024-10-17T04:00:29Z)
• Unifying Global and Near-Context Biasing in a Single Trie Pass [11.277273712268897]
  We propose an unexplored combination of an NE bias list and a word-level n-gram language model (LM)<n>We show that the proposed combination of keyword biasing and n-gram LM improves entity recognition by up to 32% relative and reduces overall WER by up to a 12% relative.
  arXiv  Detail & Related papers  (2024-09-20T13:53:37Z)
• Self-Taught Recognizer: Toward Unsupervised Adaptation for Speech Foundation Models [84.8919069953397]
  Self-TAught Recognizer (STAR) is an unsupervised adaptation framework for speech recognition systems. We show that STAR achieves an average of 13.5% relative reduction in word error rate across 14 target domains. STAR exhibits high data efficiency that only requires less than one-hour unlabeled data.
  arXiv  Detail & Related papers  (2024-05-23T04:27:11Z)
• DANCER: Entity Description Augmented Named Entity Corrector for Automatic Speech Recognition [10.844822448167935]
  We propose a Description Augmented Named entity CorrEctoR (dubbed DANCER) to mitigate phonetic confusion on automatic speech recognition (E2E ASR) transcriptions. DANCER outperforms a strong baseline, the phonetic edit-distance-based NEC model (PED-NEC), by a character error rate (CER) reduction of about 7% relatively on AISHELL-1 for named entities. More notably, when tested on Homophone that contain named entities of high phonetic confusion, DANCER offers a more pronounced CER reduction of 46% relatively over PED-NEC for named entities.
  arXiv  Detail & Related papers  (2024-03-26T12:27:32Z)
• HyPoradise: An Open Baseline for Generative Speech Recognition with Large Language Models [81.56455625624041]
  We introduce the first open-source benchmark to utilize external large language models (LLMs) for ASR error correction. The proposed benchmark contains a novel dataset, HyPoradise (HP), encompassing more than 334,000 pairs of N-best hypotheses. LLMs with reasonable prompt and its generative capability can even correct those tokens that are missing in N-best list.
  arXiv  Detail & Related papers  (2023-09-27T14:44:10Z)
• Unsupervised Syntactically Controlled Paraphrase Generation with Abstract Meaning Representations [59.10748929158525]
  Abstract Representations (AMR) can greatly improve the performance of unsupervised syntactically controlled paraphrase generation. Our proposed model, AMR-enhanced Paraphrase Generator (AMRPG), encodes the AMR graph and the constituency parses the input sentence into two disentangled semantic and syntactic embeddings. Experiments show that AMRPG generates more accurate syntactically controlled paraphrases, both quantitatively and qualitatively, compared to the existing unsupervised approaches.
  arXiv  Detail & Related papers  (2022-11-02T04:58:38Z)
• Towards Contextual Spelling Correction for Customization of End-to-end Speech Recognition Systems [27.483603895258437]
  We introduce a novel approach to do contextual biasing by adding a contextual spelling correction model on top of the end-to-end ASR system. We propose filtering algorithms to handle large-size context lists, and performance balancing mechanisms to control the biasing degree of the model. Experiments show that the proposed method achieves as much as 51% relative word error rate (WER) reduction over ASR system and outperforms traditional biasing methods.
  arXiv  Detail & Related papers  (2022-03-02T06:00:48Z)
• Joint Contextual Modeling for ASR Correction and Language Understanding [60.230013453699975]
  We propose multi-task neural approaches to perform contextual language correction on ASR outputs jointly with language understanding (LU) We show that the error rates of off the shelf ASR and following LU systems can be reduced significantly by 14% relative with joint models trained using small amounts of in-domain data.
  arXiv  Detail & Related papers  (2020-01-28T22:09:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.