arXiv 2602.188532026-03-09cs.CV

面向城市场景分割的开放词汇领域泛化

Open-Vocabulary Domain Generalization in Urban-Scene Segmentation

Department of Information Engineering and Computer Science, University of Trento（信息工程与计算机科学系，特伦托大学）；School of Computer Science and Information Engineering, Hefei University of Technology（计算机科学与信息工程学院，合肥工业大学）

Dong Zhao, Qi Zang, Nan Pu, Wenjing Li, Nicu Sebe, Zhun Zhong

发表信息：: CVPR2026

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出OVDG-SS，通过S2-Corr机制提升城市场景分割中开放词汇领域的泛化能力和鲁棒性。

摘要：

领域泛化在语义分割（DG-SS）中旨在使分割模型在未见环境中表现稳健。然而，传统DG-SS方法受限于固定已知类别的集合，限制了其在开放世界场景中的应用。最近视觉-语言模型（VLMs）的进步推动了开放词汇语义分割（OV-SS）的发展，使模型能够识别更广泛的概念。然而，这些模型仍对领域偏移敏感，在部署到未见环境中时难以保持稳健性，特别是在城市驾驶场景中尤为严重。为弥合这一差距，我们引入了面向语义分割的开放词汇领域泛化（OVDG-SS），一种新的设置，同时解决未见领域和未见类别。我们引入了首个OVDG-SS在自动驾驶中的基准，解决了一个此前未探索的问题，并覆盖了合成到现实和现实到现实的泛化，跨越多样化的未见领域和未见类别。在OVDG-SS中，我们观察到领域偏移经常扭曲预训练VLMs中的文本-图像相关性，这会阻碍OV-SS模型的性能。为解决这一挑战，我们提出了S2-Corr，一种状态空间驱动的文本-图像相关性细化机制，该机制减轻了领域引起的扭曲，并在分布变化下产生更一致的文本-图像相关性。在我们构建的基准上进行的广泛实验表明，所提出的方法在跨领域性能和效率上优于现有OV-SS方法。

英文摘要：

Domain Generalization in Semantic Segmentation (DG-SS) aims to enable segmentation models to perform robustly in unseen environments. However, conventional DG-SS methods are restricted to a fixed set of known categories, limiting their applicability in open-world scenarios. Recent progress in Vision-Language Models (VLMs) has advanced Open-Vocabulary Semantic Segmentation (OV-SS) by enabling models to recognize a broader range of concepts. Yet, these models remain sensitive to domain shifts and struggle to maintain robustness when deployed in unseen environments, a challenge that is particularly severe in urban-driving scenarios. To bridge this gap, we introduce Open-Vocabulary Domain Generalization in Semantic Segmentation (OVDG-SS), a new setting that jointly addresses unseen domains and unseen categories. We introduce the first benchmark for OVDG-SS in autonomous driving, addressing a previously unexplored problem and covering both synthetic-to-real and real-to-real generalization across diverse unseen domains and unseen categories. In OVDG-SS, we observe that domain shifts often distort text-image correlations in pre-trained VLMs, which hinders the performance of OV-SS models. To tackle this challenge, we propose S2-Corr, a state-space-driven text-image correlation refinement mechanism that mitigates domain-induced distortions and produces more consistent text-image correlations under distribution changes. Extensive experiments on our constructed benchmark demonstrate that the proposed method achieves superior cross-domain performance and efficiency compared to existing OV-SS approaches.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。