DALR: Dual-level Alignment Learning for Multimodal Sentence Representation Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: He, Kang, Ding, Yuzhe, Wang, Haining, Li, Fei, Teng, Chong, Ji, Donghong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915365907333120
author He, Kang
Ding, Yuzhe
Wang, Haining
Li, Fei
Teng, Chong
Ji, Donghong
author_facet He, Kang
Ding, Yuzhe
Wang, Haining
Li, Fei
Teng, Chong
Ji, Donghong
contents Previous multimodal sentence representation learning methods have achieved impressive performance. However, most approaches focus on aligning images and text at a coarse level, facing two critical challenges:cross-modal misalignment bias and intra-modal semantic divergence, which significantly degrade sentence representation quality. To address these challenges, we propose DALR (Dual-level Alignment Learning for Multimodal Sentence Representation). For cross-modal alignment, we propose a consistency learning module that softens negative samples and utilizes semantic similarity from an auxiliary task to achieve fine-grained cross-modal alignment. Additionally, we contend that sentence relationships go beyond binary positive-negative labels, exhibiting a more intricate ranking structure. To better capture these relationships and enhance representation quality, we integrate ranking distillation with global intra-modal alignment learning. Comprehensive experiments on semantic textual similarity (STS) and transfer (TR) tasks validate the effectiveness of our approach, consistently demonstrating its superiority over state-of-the-art baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2506_21096
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DALR: Dual-level Alignment Learning for Multimodal Sentence Representation Learning
He, Kang
Ding, Yuzhe
Wang, Haining
Li, Fei
Teng, Chong
Ji, Donghong
Computation and Language
Previous multimodal sentence representation learning methods have achieved impressive performance. However, most approaches focus on aligning images and text at a coarse level, facing two critical challenges:cross-modal misalignment bias and intra-modal semantic divergence, which significantly degrade sentence representation quality. To address these challenges, we propose DALR (Dual-level Alignment Learning for Multimodal Sentence Representation). For cross-modal alignment, we propose a consistency learning module that softens negative samples and utilizes semantic similarity from an auxiliary task to achieve fine-grained cross-modal alignment. Additionally, we contend that sentence relationships go beyond binary positive-negative labels, exhibiting a more intricate ranking structure. To better capture these relationships and enhance representation quality, we integrate ranking distillation with global intra-modal alignment learning. Comprehensive experiments on semantic textual similarity (STS) and transfer (TR) tasks validate the effectiveness of our approach, consistently demonstrating its superiority over state-of-the-art baselines.
title DALR: Dual-level Alignment Learning for Multimodal Sentence Representation Learning
topic Computation and Language
url https://arxiv.org/abs/2506.21096