Semantic Relation-Enhanced CLIP Adapter for Domain Adaptive Zero-Shot Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Jiaao, Han, Mingjie, Jiang, Jinkun, Dong, Junyu, Gong, Tao, Lan, Man
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909869736460288
author Yu, Jiaao
Han, Mingjie
Jiang, Jinkun
Dong, Junyu
Gong, Tao
Lan, Man
author_facet Yu, Jiaao
Han, Mingjie
Jiang, Jinkun
Dong, Junyu
Gong, Tao
Lan, Man
contents The high cost of data annotation has spurred research on training deep learning models in data-limited scenarios. Existing paradigms, however, fail to balance cross-domain transfer and cross-category generalization, giving rise to the demand for Domain-Adaptive Zero-Shot Learning (DAZSL). Although vision-language models (e.g., CLIP) have inherent advantages in the DAZSL field, current studies do not fully exploit their potential. Applying CLIP to DAZSL faces two core challenges: inefficient cross-category knowledge transfer due to the lack of semantic relation guidance, and degraded cross-modal alignment during target domain fine-tuning. To address these issues, we propose a Semantic Relation-Enhanced CLIP (SRE-CLIP) Adapter framework, integrating a Semantic Relation Structure Loss and a Cross-Modal Alignment Retention Strategy. As the first CLIP-based DAZSL method, SRE-CLIP achieves state-of-the-art performance on the I2AwA and I2WebV benchmarks, significantly outperforming existing approaches.
format Preprint
id arxiv_https___arxiv_org_abs_2510_21808
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Semantic Relation-Enhanced CLIP Adapter for Domain Adaptive Zero-Shot Learning
Yu, Jiaao
Han, Mingjie
Jiang, Jinkun
Dong, Junyu
Gong, Tao
Lan, Man
Computer Vision and Pattern Recognition
Artificial Intelligence
The high cost of data annotation has spurred research on training deep learning models in data-limited scenarios. Existing paradigms, however, fail to balance cross-domain transfer and cross-category generalization, giving rise to the demand for Domain-Adaptive Zero-Shot Learning (DAZSL). Although vision-language models (e.g., CLIP) have inherent advantages in the DAZSL field, current studies do not fully exploit their potential. Applying CLIP to DAZSL faces two core challenges: inefficient cross-category knowledge transfer due to the lack of semantic relation guidance, and degraded cross-modal alignment during target domain fine-tuning. To address these issues, we propose a Semantic Relation-Enhanced CLIP (SRE-CLIP) Adapter framework, integrating a Semantic Relation Structure Loss and a Cross-Modal Alignment Retention Strategy. As the first CLIP-based DAZSL method, SRE-CLIP achieves state-of-the-art performance on the I2AwA and I2WebV benchmarks, significantly outperforming existing approaches.
title Semantic Relation-Enhanced CLIP Adapter for Domain Adaptive Zero-Shot Learning
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2510.21808