@inproceedings{guo-etal-2025-parameter,
title = "A Parameter-Efficient and Fine-Grained Prompt Learning for Vision-Language Models",
author = "Guo, Yongbin and
Li, Shuzhen and
Liu, Zhulin and
Zhang, Tong and
Chen, C.L.Philip",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.1514/",
doi = "10.18653/v1/2025.acl-long.1514",
pages = "31346--31359",
ISBN = "979-8-89176-251-0",
abstract = "Current vision-language models (VLMs) understand complex vision-text tasks by extracting overall semantic information from large-scale cross-modal associations. However, extracting from large-scale cross-modal associations often smooths out semantic details and requires large computations, limiting multimodal fine-grained understanding performance and efficiency. To address this issue, this paper proposes a detail-oriented prompt learning (DoPL) method for vision-language models to implement fine-grained multi-modal semantic alignment with merely 0.25M trainable parameters. According to the low-entropy information concentration theory, DoPL explores shared interest tokens from text-vision correlations and transforms them into alignment weights to enhance text prompt and vision prompt via detail-oriented prompt generation. It effectively guides the current frozen layer to extract fine-grained text-vision alignment cues. Furthermore, DoPL constructs detail-oriented prompt generation for each frozen layer to implement layer-by-layer localization of fine-grained semantic alignment, achieving precise understanding in complex vision-text tasks. DoPL performs well in parameter-efficient fine-grained semantic alignment with only 0.12{\%} tunable parameters for vision-language models. The state-of-the-art results over the previous parameter-efficient fine-tuning methods and full fine-tuning approaches on six benchmarks demonstrate the effectiveness and efficiency of DoPL in complex multi-modal tasks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="guo-etal-2025-parameter">
<titleInfo>
<title>A Parameter-Efficient and Fine-Grained Prompt Learning for Vision-Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yongbin</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shuzhen</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhulin</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">C.L.Philip</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>Current vision-language models (VLMs) understand complex vision-text tasks by extracting overall semantic information from large-scale cross-modal associations. However, extracting from large-scale cross-modal associations often smooths out semantic details and requires large computations, limiting multimodal fine-grained understanding performance and efficiency. To address this issue, this paper proposes a detail-oriented prompt learning (DoPL) method for vision-language models to implement fine-grained multi-modal semantic alignment with merely 0.25M trainable parameters. According to the low-entropy information concentration theory, DoPL explores shared interest tokens from text-vision correlations and transforms them into alignment weights to enhance text prompt and vision prompt via detail-oriented prompt generation. It effectively guides the current frozen layer to extract fine-grained text-vision alignment cues. Furthermore, DoPL constructs detail-oriented prompt generation for each frozen layer to implement layer-by-layer localization of fine-grained semantic alignment, achieving precise understanding in complex vision-text tasks. DoPL performs well in parameter-efficient fine-grained semantic alignment with only 0.12% tunable parameters for vision-language models. The state-of-the-art results over the previous parameter-efficient fine-tuning methods and full fine-tuning approaches on six benchmarks demonstrate the effectiveness and efficiency of DoPL in complex multi-modal tasks.</abstract>
<identifier type="citekey">guo-etal-2025-parameter</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.1514</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.1514/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>31346</start>
<end>31359</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A Parameter-Efficient and Fine-Grained Prompt Learning for Vision-Language Models
%A Guo, Yongbin
%A Li, Shuzhen
%A Liu, Zhulin
%A Zhang, Tong
%A Chen, C.L.Philip
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F guo-etal-2025-parameter
%X Current vision-language models (VLMs) understand complex vision-text tasks by extracting overall semantic information from large-scale cross-modal associations. However, extracting from large-scale cross-modal associations often smooths out semantic details and requires large computations, limiting multimodal fine-grained understanding performance and efficiency. To address this issue, this paper proposes a detail-oriented prompt learning (DoPL) method for vision-language models to implement fine-grained multi-modal semantic alignment with merely 0.25M trainable parameters. According to the low-entropy information concentration theory, DoPL explores shared interest tokens from text-vision correlations and transforms them into alignment weights to enhance text prompt and vision prompt via detail-oriented prompt generation. It effectively guides the current frozen layer to extract fine-grained text-vision alignment cues. Furthermore, DoPL constructs detail-oriented prompt generation for each frozen layer to implement layer-by-layer localization of fine-grained semantic alignment, achieving precise understanding in complex vision-text tasks. DoPL performs well in parameter-efficient fine-grained semantic alignment with only 0.12% tunable parameters for vision-language models. The state-of-the-art results over the previous parameter-efficient fine-tuning methods and full fine-tuning approaches on six benchmarks demonstrate the effectiveness and efficiency of DoPL in complex multi-modal tasks.
%R 10.18653/v1/2025.acl-long.1514
%U https://aclanthology.org/2025.acl-long.1514/
%U https://doi.org/10.18653/v1/2025.acl-long.1514
%P 31346-31359
Markdown (Informal)
[A Parameter-Efficient and Fine-Grained Prompt Learning for Vision-Language Models](https://aclanthology.org/2025.acl-long.1514/) (Guo et al., ACL 2025)
ACL