大模型
arxiv.org ·
通过文本去噪实现基于LLM的ASR的纯文本适配
该论文针对基于 LLM 的自动语音识别（ASR）系统在仅有文本数据时进行新领域适配的难题：直接对 LLM 做目标域文本微调常会破坏 projector 学到的语音-文本模态对齐。作者提出把纯文本适配建模为文本去噪任务，训练 LLM 从含噪输入恢复干净转写，从而在适配目标域的同时保持跨模态对齐。方法轻量，无需改动架构或增加参数。在两个数据集上的评测显示最高 22.1% 的相对提升，作者称优于近期纯文本适配方法。
AI推荐理由 arXiv 论文（v4，2026-09-22 修订），作者与摘要明确，属可验证的研究成果。核心新信息：提出将基于 LLM 的 ASR 的纯文本域适配重构为文本去噪任务，训练 LLM 从含噪输入恢复干净转写，从而在适配目标域的同时保留语音-文本跨模态对齐；无需架构改动或额外参数，轻量。摘要称在两个数据集上取得最高 22.1% 相对提升，并优于近期纯文本适配方法。局限：仅摘要级证据，未提供数据集名称、基线细节与复现条件，22.1% 为作者自报相对提升，需看正文验证。
78
正文
Computer Science > Sound
arXiv:2601.20900 (cs)
[Submitted on 28 Jan 2026 (v1), last revised 22 Sep 2026 (this version, v4)]
Title:Text-only adaptation in LLM-based ASR through text denoising
Authors:Andrés Carofilis, Sergio Burdisso, Esaú Villatoro-Tello, Shashi Kumar, Kadri Hacioglu, Srikanth Madikeri, Pradeep Rangappa, Manjunath K E, Petr Motlicek, Shankar Venkatesan, Andreas Stolcke
View a PDF of the paper titled Text-only adaptation in LLM-based ASR through text denoising, by Andr\'es Carofilis and 10 other authors
View PDF
HTML (experimental)
Abstract:Adapting large language model (LLM)-based automatic speech recognition (ASR) systems to new domains using text-only data is a significant yet underexplored challenge. Standard fine-tuning of the LLM on the target domain text often disrupts the critical alignment between the speech and text modality learned by the projector, degrading performance. We introduce a novel text-only adaptation method that frames this process as a text denoising task. Our approach trains the LLM to recover clean transcripts from noisy inputs. This process effectively adapts the model to a target domain while preserving cross-modal alignment. Our solution is lightweight, requiring no architectural changes or additional parameters. Extensive evaluation on two datasets demonstrates up to 22.1% relative improvement, outperforming recent state-of-the-art text-only adaptation methods.
Comments:
Subjects:
Sound (cs.SD); Computation and Language (cs.CL); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
Cite as:
arXiv:2601.20900 [cs.SD]
(or
arXiv:2601.20900v4 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2601.20900
Focus to learn more
arXiv-issued DOI via DataCite
Submission history
From: Sergio Burdisso [view email]
[v1]
Wed, 28 Jan 2026 10:18:23 UTC (234 KB)
[v2]
Fri, 30 Jan 2026 08:53:08 UTC (234 KB)
[v3]
Thu, 12 Mar 2026 09:27:02 UTC (266 KB)
[v4]
Tue, 22 Sep 2026 08:06:44 UTC (236 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Text-only adaptation in LLM-based ASR through text denoising, by Andr\'es Carofilis and 10 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.SD
new
|
recent
| 2026-01
Change to browse by:
cs
cs.CL
cs.LG
eess
eess.AS
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? |
Disable MathJax (What is MathJax?)