EReLiFM: Evidential Reliability-Aware Residual Flow Meta-Learning for Open-Set Domain Generalization under Noisy Labels

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Peng, Kunyu, Wen, Di, Yang, Kailun, Fu, Jia, Chen, Yufan, Liu, Ruiping, Wu, Jiamin, Zheng, Junwei, Sarfraz, M. Saquib, Van Gool, Luc, Paudel, Danda Pani, Stiefelhagen, Rainer
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912647764508672
author Peng, Kunyu
Wen, Di
Yang, Kailun
Fu, Jia
Chen, Yufan
Liu, Ruiping
Wu, Jiamin
Zheng, Junwei
Sarfraz, M. Saquib
Van Gool, Luc
Paudel, Danda Pani
Stiefelhagen, Rainer
author_facet Peng, Kunyu
Wen, Di
Yang, Kailun
Fu, Jia
Chen, Yufan
Liu, Ruiping
Wu, Jiamin
Zheng, Junwei
Sarfraz, M. Saquib
Van Gool, Luc
Paudel, Danda Pani
Stiefelhagen, Rainer
contents Open-Set Domain Generalization (OSDG) aims to enable deep learning models to recognize unseen categories in new domains, which is crucial for real-world applications. Label noise hinders open-set domain generalization by corrupting source-domain knowledge, making it harder to recognize known classes and reject unseen ones. While existing methods address OSDG under Noisy Labels (OSDG-NL) using hyperbolic prototype-guided meta-learning, they struggle to bridge domain gaps, especially with limited clean labeled data. In this paper, we propose Evidential Reliability-Aware Residual Flow Meta-Learning (EReLiFM). We first introduce an unsupervised two-stage evidential loss clustering method to promote label reliability awareness. Then, we propose a residual flow matching mechanism that models structured domain- and category-conditioned residuals, enabling diverse and uncertainty-aware transfer paths beyond interpolation-based augmentation. During this meta-learning process, the model is optimized such that the update direction on the clean set maximizes the loss decrease on the noisy set, using pseudo labels derived from the most confident predicted class for supervision. Experimental results show that EReLiFM outperforms existing methods on OSDG-NL, achieving state-of-the-art performance. The source code is available at https://github.com/KPeng9510/ERELIFM.
format Preprint
id arxiv_https___arxiv_org_abs_2510_12687
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle EReLiFM: Evidential Reliability-Aware Residual Flow Meta-Learning for Open-Set Domain Generalization under Noisy Labels
Peng, Kunyu
Wen, Di
Yang, Kailun
Fu, Jia
Chen, Yufan
Liu, Ruiping
Wu, Jiamin
Zheng, Junwei
Sarfraz, M. Saquib
Van Gool, Luc
Paudel, Danda Pani
Stiefelhagen, Rainer
Computer Vision and Pattern Recognition
Machine Learning
Robotics
Open-Set Domain Generalization (OSDG) aims to enable deep learning models to recognize unseen categories in new domains, which is crucial for real-world applications. Label noise hinders open-set domain generalization by corrupting source-domain knowledge, making it harder to recognize known classes and reject unseen ones. While existing methods address OSDG under Noisy Labels (OSDG-NL) using hyperbolic prototype-guided meta-learning, they struggle to bridge domain gaps, especially with limited clean labeled data. In this paper, we propose Evidential Reliability-Aware Residual Flow Meta-Learning (EReLiFM). We first introduce an unsupervised two-stage evidential loss clustering method to promote label reliability awareness. Then, we propose a residual flow matching mechanism that models structured domain- and category-conditioned residuals, enabling diverse and uncertainty-aware transfer paths beyond interpolation-based augmentation. During this meta-learning process, the model is optimized such that the update direction on the clean set maximizes the loss decrease on the noisy set, using pseudo labels derived from the most confident predicted class for supervision. Experimental results show that EReLiFM outperforms existing methods on OSDG-NL, achieving state-of-the-art performance. The source code is available at https://github.com/KPeng9510/ERELIFM.
title EReLiFM: Evidential Reliability-Aware Residual Flow Meta-Learning for Open-Set Domain Generalization under Noisy Labels
topic Computer Vision and Pattern Recognition
Machine Learning
Robotics
url https://arxiv.org/abs/2510.12687