CLIPin: A Non-contrastive Plug-in to CLIP for Multimodal Semantic Alignment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yang, Shengzhu, Du, Jiawei, Lu, Shuai, Zhang, Weihang, Wang, Ningli, Li, Huiqi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916968995487744
author Yang, Shengzhu
Du, Jiawei
Lu, Shuai
Zhang, Weihang
Wang, Ningli
Li, Huiqi
author_facet Yang, Shengzhu
Du, Jiawei
Lu, Shuai
Zhang, Weihang
Wang, Ningli
Li, Huiqi
contents Large-scale natural image-text datasets, especially those automatically collected from the web, often suffer from loose semantic alignment due to weak supervision, while medical datasets tend to have high cross-modal correlation but low content diversity. These properties pose a common challenge for contrastive language-image pretraining (CLIP): they hinder the model's ability to learn robust and generalizable representations. In this work, we propose CLIPin, a unified non-contrastive plug-in that can be seamlessly integrated into CLIP-style architectures to improve multimodal semantic alignment, providing stronger supervision and enhancing alignment robustness. Furthermore, two shared pre-projectors are designed for image and text modalities respectively to facilitate the integration of contrastive and non-contrastive learning in a parameter-compromise manner. Extensive experiments on diverse downstream tasks demonstrate the effectiveness and generality of CLIPin as a plug-and-play component compatible with various contrastive frameworks. Code is available at https://github.com/T6Yang/CLIPin.
format Preprint
id arxiv_https___arxiv_org_abs_2508_06434
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CLIPin: A Non-contrastive Plug-in to CLIP for Multimodal Semantic Alignment
Yang, Shengzhu
Du, Jiawei
Lu, Shuai
Zhang, Weihang
Wang, Ningli
Li, Huiqi
Computer Vision and Pattern Recognition
Artificial Intelligence
Large-scale natural image-text datasets, especially those automatically collected from the web, often suffer from loose semantic alignment due to weak supervision, while medical datasets tend to have high cross-modal correlation but low content diversity. These properties pose a common challenge for contrastive language-image pretraining (CLIP): they hinder the model's ability to learn robust and generalizable representations. In this work, we propose CLIPin, a unified non-contrastive plug-in that can be seamlessly integrated into CLIP-style architectures to improve multimodal semantic alignment, providing stronger supervision and enhancing alignment robustness. Furthermore, two shared pre-projectors are designed for image and text modalities respectively to facilitate the integration of contrastive and non-contrastive learning in a parameter-compromise manner. Extensive experiments on diverse downstream tasks demonstrate the effectiveness and generality of CLIPin as a plug-and-play component compatible with various contrastive frameworks. Code is available at https://github.com/T6Yang/CLIPin.
title CLIPin: A Non-contrastive Plug-in to CLIP for Multimodal Semantic Alignment
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2508.06434