ICRL: independent causality representation learning for domain generalization

Abstract Domain generalization (DG) addresses the challenge of out-of-distribution (OOD) data; however, the reliance on statistical correlations during model development often introduces shortcut learning problems. Current approaches to mitigating these issues commonly involve the integration of cau...

Full description

Saved in:
Bibliographic Details
Main Authors: Liwen Xu, Yuxuan Shao
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-96357-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849726371205480448
author Liwen Xu
Yuxuan Shao
author_facet Liwen Xu
Yuxuan Shao
author_sort Liwen Xu
collection DOAJ
description Abstract Domain generalization (DG) addresses the challenge of out-of-distribution (OOD) data; however, the reliance on statistical correlations during model development often introduces shortcut learning problems. Current approaches to mitigating these issues commonly involve the integration of causal inference, formalizing DG problems through a general structural causal model. Nevertheless, ensuring the independence of features when incorporating causal models is often overlooked, leading to spurious causal relationships. In this work, we design three independent feature modules using GAN variants (GAN, WGAN, and WGAN-GP) and select the best-performing WGAN module to integrate into the existing causal model framework, constructing an independent causal relationship learning (ICRL) model. Extensive experiments on widely used datasets demonstrate that our model, with independent causal representations, outperforms the original model in both performance and efficiency, thereby validating the effectiveness of our proposed approach. The code for ICRL can be accessed at: https://github.com/22Shao/ICRL.git .
format Article
id doaj-art-c06cce0df7e848da8a26514ea7477ee1
institution DOAJ
issn 2045-2322
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-c06cce0df7e848da8a26514ea7477ee12025-08-20T03:10:11ZengNature PortfolioScientific Reports2045-23222025-04-0115111510.1038/s41598-025-96357-0ICRL: independent causality representation learning for domain generalizationLiwen Xu0Yuxuan Shao1College of Science, North China University of TechnologyCollege of Science, North China University of TechnologyAbstract Domain generalization (DG) addresses the challenge of out-of-distribution (OOD) data; however, the reliance on statistical correlations during model development often introduces shortcut learning problems. Current approaches to mitigating these issues commonly involve the integration of causal inference, formalizing DG problems through a general structural causal model. Nevertheless, ensuring the independence of features when incorporating causal models is often overlooked, leading to spurious causal relationships. In this work, we design three independent feature modules using GAN variants (GAN, WGAN, and WGAN-GP) and select the best-performing WGAN module to integrate into the existing causal model framework, constructing an independent causal relationship learning (ICRL) model. Extensive experiments on widely used datasets demonstrate that our model, with independent causal representations, outperforms the original model in both performance and efficiency, thereby validating the effectiveness of our proposed approach. The code for ICRL can be accessed at: https://github.com/22Shao/ICRL.git .https://doi.org/10.1038/s41598-025-96357-0IndependenceCausal inferenceDomain generalization
spellingShingle Liwen Xu
Yuxuan Shao
ICRL: independent causality representation learning for domain generalization
Scientific Reports
Independence
Causal inference
Domain generalization
title ICRL: independent causality representation learning for domain generalization
title_full ICRL: independent causality representation learning for domain generalization
title_fullStr ICRL: independent causality representation learning for domain generalization
title_full_unstemmed ICRL: independent causality representation learning for domain generalization
title_short ICRL: independent causality representation learning for domain generalization
title_sort icrl independent causality representation learning for domain generalization
topic Independence
Causal inference
Domain generalization
url https://doi.org/10.1038/s41598-025-96357-0
work_keys_str_mv AT liwenxu icrlindependentcausalityrepresentationlearningfordomaingeneralization
AT yuxuanshao icrlindependentcausalityrepresentationlearningfordomaingeneralization