{"type":"ScholarlyArticle","title":"MarCon: Max-Margin Contrastive Learning for Imbalanced Domain Adaptation Semantic Segmentation","authors":[{"name":"Yibo Wang","slug":"yibo-wang","given":"Yibo","family":"Wang","affiliation":""},{"name":"Ruikang Xu","slug":"ruikang-xu","given":"Ruikang","family":"Xu","affiliation":""},{"name":"Guangcheng Zhu","slug":"guangcheng-zhu","given":"Guangcheng","family":"Zhu","affiliation":""},{"name":"Cheng Peng","slug":"cheng-peng","given":"Cheng","family":"Peng","affiliation":""},{"name":"Haobo Wang","slug":"haobo-wang","given":"Haobo","family":"Wang","affiliation":""},{"name":"Runze Wu","slug":"runze-wu","given":"Runze","family":"Wu","affiliation":""},{"name":"Minmin Lin","slug":"minmin-lin","given":"Minmin","family":"Lin","affiliation":""},{"name":"Changjie Fan","slug":"changjie-fan","given":"Changjie","family":"Fan","affiliation":""}],"abstract":"Unsupervised Domain Adaptation for Semantic Segmentation (UDA-SS) has seen significant progress in recent years. Existing UDA-SS approaches mostly adopt a pseudo-labeling schema to adapt model in the target domain, but they often overlook the inherent long-tailed data distribution in segmentation. We find that such scarce tail samples can lead to representation collapse for tail classes and further hinder the quality of pseudo-labels. To address these challenges, we propose MarCon, a framework that mitigates the long-tailed problem from both feature representation and pseudo-labeling perspectives. Specifically, to explicitly learn a Maximum-Margin Distribution, we derive a reformulated pixel-level contrastive learning objective by modeling feature distributions with the von Mises-Fisher (vMF) distribution. It enforces strict margins to enhance intra-class compactness and inter-class separability, preventing tail classes from being overwhelmed by head classes. Furthermore, to mitigate label noise, we introduce a Reliability-aware Filter (RaF) based on the vMF-derived metrics, which performs adaptive class-wise pixel reliability assessment to identify unreliable pixels and attenuate their contribution to model training, thereby mitigating confirmation bias. Extensive experiments on GTA → Cityscapes and SYNTHIA → Cityscapes demonstrate that MarCon consistently outperforms current leading methods across various transformer-based architectures.","doi":"10.24963/ijcai.2026/556","url":"https://ijcai.org/proceedings/2026/556","pdf":"https://ijcai.org/proceedings/2026/0556.pdf","markdown":"https://ijcai.org/proceedings/2026/556.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/556","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Machine Learning","track":"Main Track","number":556,"pages":{"first":4994,"last":5002},"keywords":["Machine Learning: Weakly supervised learning"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}