Journal
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
Volume 45, Issue 7, Pages 8827-8844Publisher
IEEE COMPUTER SOC
DOI: 10.1109/TPAMI.2022.3233584
Keywords
Cross-image semantic consistency; label rectifying; semantic segmentation; semi-supervised learning
Ask authors/readers for more resources
This paper proposes a semi-supervised semantic segmentation method that learns a segmentation model using limited labeled images and abundant unlabeled images. The approach focuses on generating reliable pseudo labels based on the confidence scores of unlabeled images, but overlooks the use of labeled images. The proposed CISC-R method leverages labeled images to refine the pseudo labels and enhance the performance of semi-supervised semantic segmentation.
Semi-supervised semantic segmentation aims to learn a semantic segmentation model via limited labeled images and adequate unlabeled images. The key to this task is generating reliable pseudo labels for unlabeled images. Existing methods mainly focus on producing reliable pseudo labels based on the confidence scores of unlabeled images while largely ignoring the use of labeled images with accurate annotations. In this paper, we propose a Cross-Image Semantic Consistency guided Rectifying (CISC-R) approach for semi-supervised semantic segmentation, which explicitly leverages the labeled images to rectify the generated pseudo labels. Our CISC-R is inspired by the fact that images belonging to the same class have a high pixel-level correspondence. Specifically, given an unlabeled image and its initial pseudo labels, we first query a guiding labeled image that shares the same semantic information with the unlabeled image. Then, we estimate the pixel-level similarity between the unlabeled image and the queried labeled image to form a CISC map, which guides us to achieve a reliable pixel-level rectification for the pseudo labels. Extensive experiments on the PASCALVOC 2012, Cityscapes, and COCO datasets demonstrate that the proposed CISC-R can significantly improve the quality of the pseudo labels and outperform the state-of-the-art methods. Code is available at https://github.com/Luffy03/CISC-R.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available