SemiDAViL：面向语义分割的视觉-语言引导半监督域自适应
SemiDAViL: Semi-supervised Domain Adaptation with Vision-Language Guidance for Semantic Segmentation
AI总结：
本文提出首个语言引导的半监督域自适应语义分割框架SemiDAViL，利用视觉-语言模型增强跨域特征判别，并通过类别平衡分割损失缓解长尾类别不平衡，在多个域自适应场景中显著超越现有最先进方法。
摘要：
域自适应（DA）与半监督学习（SSL）在半监督域自适应（SSDA）中交汇，其目标是通过结合有限的目标域标注样本和大量未标注目标域数据，将知识从源域迁移到目标域。尽管这一思路直观，但在语义分割中简单融合DA与SSL并非最优，原因主要有两点：（1）以往方法虽然能够学习到良好的分割边界，但由于监督有限，容易混淆视觉外观相似的类别；（2）偏斜且不平衡的训练数据分布倾向于源域表示学习，阻碍了对尾部类别有限信息的挖掘。语言引导可以充当关键的语义桥梁，通过利用预训练语言模型中编码的丰富语义关系来增强跨域特征表示，从而促进鲁棒的类别判别并缓解视觉歧义。因此，本文首次提出面向语义分割的语言引导SSDA设置。具体而言，我们利用视觉-语言模型（VLMs）固有的语义泛化能力，在SSDA范式内建立一个协同框架。为应对长尾分布中固有的类别不平衡挑战，我们引入类别平衡的分割损失公式，有效正则化学习过程。通过在多种域自适应场景下的大量实验，我们的方法相比当前最先进（SoTA）方法展现出显著的性能提升。代码已开源：\href{https://github.com/hritam-98/SemiDAViL}{GitHub}。
英文摘要：
Domain Adaptation (DA) and Semi-supervised Learning (SSL) converge in Semi-supervised Domain Adaptation (SSDA), where the objective is to transfer knowledge from a source domain to a target domain using a combination of limited labeled target samples and abundant unlabeled target data. Although intuitive, a simple amalgamation of DA and SSL is suboptimal in semantic segmentation due to two major reasons: (1) previous methods, while able to learn good segmentation boundaries, are prone to confuse classes with similar visual appearance due to limited supervision; and (2) skewed and imbalanced training data distribution preferring source representation learning whereas impeding from exploring limited information about tailed classes. Language guidance can serve as a pivotal semantic bridge, facilitating robust class discrimination and mitigating visual ambiguities by leveraging the rich semantic relationships encoded in pre-trained language models to enhance feature representations across domains. Therefore, we propose the first language-guided SSDA setting for semantic segmentation in this work. Specifically, we harness the semantic generalization capabilities inherent in vision-language models (VLMs) to establish a synergistic framework within the SSDA paradigm. To address the inherent class-imbalance challenges in long-tailed distributions, we introduce class-balanced segmentation loss formulations that effectively regularize the learning process. Through extensive experimentation across diverse domain adaptation scenarios, our approach demonstrates substantial performance improvements over contemporary state-of-the-art (SoTA) methodologies. Code is available: \href{https://github.com/hritam-98/SemiDAViL}{GitHub}.
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。
