arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.04036cs.CV
收藏

超越像素:基于多尺度补丁多标签分类器的半监督语义分割

Beyond Pixels: Semi-Supervised Semantic Segmentation with a Multi-scale Patch-based Multi-Label Classifier

Stonybrook University(石溪大学);University of North Carolina at Charlotte(北卡罗来纳大学夏洛特分校);EPFL(洛桑联邦理工学院)

Prantik Howlader, Srijan Das, Hieu Le, Dimitris Samaras

AI总结:

本文提出多尺度补丁多标签分类器(MPMC),作为即插即用模块,通过补丁级监督和自适应伪标签权重,提升现有半监督语义分割框架性能,在三个数据集上显著改善分割结果。

摘要:

整合像素上下文信息对于精确分割至关重要。在本文中,我们展示了一种有效整合上下文信息的方式,即通过基于补丁的分类器。该补丁分类器被训练用于识别图像区域内存在的类别,这有助于消除干扰物并增强对小物体分割区域的分类。具体而言,我们引入了多尺度补丁多标签分类器(MPMC),这是一种新颖的即插即用模块,专为现有的半监督分割(SSS)框架设计。MPMC提供补丁级监督,使得能够区分补丁内不同类别的像素区域。此外,MPMC学习一个自适应的伪标签权重,利用补丁级分类来减轻教师网络对学生网络产生的噪声伪标签监督的影响。这个轻量级模块可以集成到任何SSS框架中,显著提升其性能。我们通过将所提出的MPMC集成到四种SSS方法中,并在两个自然图像和一个医学分割数据集上改进它们,证明了其有效性,显著改善了所有三个数据集上基线方法的分割结果。

英文摘要:

Incorporating pixel contextual information is critical for accurate segmentation. In this paper, we show that an effective way to incorporate contextual information is through a patch-based classifier. This patch classifier is trained to identify classes present within an image region, which facilitates the elimination of distractors and enhances the classification of small object segments. Specifically, we introduce Multi-scale Patch-based Multi-label Classifier (MPMC), a novel plug-in module designed for existing semi-supervised segmentation (SSS) frameworks. MPMC offers patch-level supervision, enabling the discrimination of pixel regions of different classes within a patch. Furthermore, MPMC learns an adaptive pseudo-label weight, using patch-level classification to alleviate the impact of the teacher's noisy pseudo-label supervision the student. This lightweight module can be integrated into any SSS framework, significantly enhancing their performance. We demonstrate the efficacy of our proposed MPMC by integrating it into four SSS methodologies and improving them across two natural image and one medical segmentation dataset, notably improving the segmentation results of the baselines across all the three datasets.

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。

↑