@inproceedings{cao-etal-2024-rap,
title = "{RAP}: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter",
author = "Cao, Meng and
Tang, Haoran and
Huang, Jinfa and
Jin, Peng and
Zhang, Can and
Liu, Ruyang and
Chen, Long and
Liang, Xiaodan and
Yuan, Li and
Li, Ge",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2024",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-acl.427",
doi = "10.18653/v1/2024.findings-acl.427",
pages = "7160--7174",
abstract = "Text-Video Retrieval (TVR) aims to align relevant video content with natural language queries. To date, most of the state-of-the-art TVR methods learn image-to-video transfer learning based on the large-scale pre-trained vision-language models (e.g., CLIP). However, fully fine-tuning these pre-trained models for TVR incurs prohibitively expensive computation cost. To this end, we propose to conduct efficient text-video Retrieval with a salient-and-correlated AdaPter (RAP), i.e., fine-tuning the pre-trained model with a few parameterized layers. To accommodate the text-video scenario, we equip our RAP with two indispensable characteristics including temporal sparsity and correlation. Specifically, we propose a low-rank modulation module to refine the per-image features from frozen CLIP backbone, which accentuates silent frames within the video features while alleviating temporal redundancy. Besides, we introduce an asynchronous self-attention mechanism which firstly selects top responsive visual patch and augments the correlation modeling between them with learnable temporal and patch offsets. Extensive experiments on four TVR datasets demonstrate that our RAP achieves superior or comparable performance compared to the fully fine-tuned counterpart and other parameter-efficient finetuning methods.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cao-etal-2024-rap">
<titleInfo>
<title>RAP: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter</title>
</titleInfo>
<name type="personal">
<namePart type="given">Meng</namePart>
<namePart type="family">Cao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haoran</namePart>
<namePart type="family">Tang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jinfa</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Peng</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Can</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruyang</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Long</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaodan</namePart>
<namePart type="family">Liang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Li</namePart>
<namePart type="family">Yuan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ge</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lun-Wei</namePart>
<namePart type="family">Ku</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Martins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Srikumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Text-Video Retrieval (TVR) aims to align relevant video content with natural language queries. To date, most of the state-of-the-art TVR methods learn image-to-video transfer learning based on the large-scale pre-trained vision-language models (e.g., CLIP). However, fully fine-tuning these pre-trained models for TVR incurs prohibitively expensive computation cost. To this end, we propose to conduct efficient text-video Retrieval with a salient-and-correlated AdaPter (RAP), i.e., fine-tuning the pre-trained model with a few parameterized layers. To accommodate the text-video scenario, we equip our RAP with two indispensable characteristics including temporal sparsity and correlation. Specifically, we propose a low-rank modulation module to refine the per-image features from frozen CLIP backbone, which accentuates silent frames within the video features while alleviating temporal redundancy. Besides, we introduce an asynchronous self-attention mechanism which firstly selects top responsive visual patch and augments the correlation modeling between them with learnable temporal and patch offsets. Extensive experiments on four TVR datasets demonstrate that our RAP achieves superior or comparable performance compared to the fully fine-tuned counterpart and other parameter-efficient finetuning methods.</abstract>
<identifier type="citekey">cao-etal-2024-rap</identifier>
<identifier type="doi">10.18653/v1/2024.findings-acl.427</identifier>
<location>
<url>https://aclanthology.org/2024.findings-acl.427</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>7160</start>
<end>7174</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T RAP: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter
%A Cao, Meng
%A Tang, Haoran
%A Huang, Jinfa
%A Jin, Peng
%A Zhang, Can
%A Liu, Ruyang
%A Chen, Long
%A Liang, Xiaodan
%A Yuan, Li
%A Li, Ge
%Y Ku, Lun-Wei
%Y Martins, Andre
%Y Srikumar, Vivek
%S Findings of the Association for Computational Linguistics: ACL 2024
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F cao-etal-2024-rap
%X Text-Video Retrieval (TVR) aims to align relevant video content with natural language queries. To date, most of the state-of-the-art TVR methods learn image-to-video transfer learning based on the large-scale pre-trained vision-language models (e.g., CLIP). However, fully fine-tuning these pre-trained models for TVR incurs prohibitively expensive computation cost. To this end, we propose to conduct efficient text-video Retrieval with a salient-and-correlated AdaPter (RAP), i.e., fine-tuning the pre-trained model with a few parameterized layers. To accommodate the text-video scenario, we equip our RAP with two indispensable characteristics including temporal sparsity and correlation. Specifically, we propose a low-rank modulation module to refine the per-image features from frozen CLIP backbone, which accentuates silent frames within the video features while alleviating temporal redundancy. Besides, we introduce an asynchronous self-attention mechanism which firstly selects top responsive visual patch and augments the correlation modeling between them with learnable temporal and patch offsets. Extensive experiments on four TVR datasets demonstrate that our RAP achieves superior or comparable performance compared to the fully fine-tuned counterpart and other parameter-efficient finetuning methods.
%R 10.18653/v1/2024.findings-acl.427
%U https://aclanthology.org/2024.findings-acl.427
%U https://doi.org/10.18653/v1/2024.findings-acl.427
%P 7160-7174
Markdown (Informal)
[RAP: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter](https://aclanthology.org/2024.findings-acl.427) (Cao et al., Findings 2024)
ACL
- Meng Cao, Haoran Tang, Jinfa Huang, Peng Jin, Can Zhang, Ruyang Liu, Long Chen, Xiaodan Liang, Li Yuan, and Ge Li. 2024. RAP: Efficient Text-Video Retrieval with Sparse-and-Correlated Adapter. In Findings of the Association for Computational Linguistics: ACL 2024, pages 7160–7174, Bangkok, Thailand. Association for Computational Linguistics.