Pixel Exclusion: Uncertainty-aware Boundary Discovery for Active Cross-Domain Semantic Segmentation
Paper
Paper/Presentation Title | Pixel Exclusion: Uncertainty-aware Boundary Discovery for Active Cross-Domain Semantic Segmentation |
---|---|
Presentation Type | Paper |
Authors | You, Fuming, Li, Jingjing, Chen, Zhi and Zhu, Lei |
Journal or Proceedings Title | Proceedings of the 30th ACM International Conference on Multimedia (MM '22) |
Journal Citation | pp. 1866-1874 |
Number of Pages | 9 |
Year | 2022 |
Publisher | Association for Computing Machinery (ACM) |
Place of Publication | United States |
ISBN | 9781450392037 |
Digital Object Identifier (DOI) | https://doi.org/10.1145/3503161.3548079 |
Web Address (URL) of Paper | https://dl.acm.org/doi/10.1145/3503161.3548079 |
Web Address (URL) of Conference Proceedings | https://dl.acm.org/doi/proceedings/10.1145/3503161 |
Conference/Event | 30th ACM International Conference on Multimedia (MM '22) |
Event Details | 30th ACM International Conference on Multimedia (MM '22) Parent ACM International Conference on Multimedia Delivery In person Event Date 10 to end of 14 Oct 2022 Event Location Lisbon, Portugal Event Web Address (URL) |
Abstract | Unsupervised Domain Adaptation (UDA) has been shown to alleviate the heavy annotations for semantic segmentation. Recently, numerous self-training approaches are proposed to address the challenging cross-domain semantic segmentation problem. However, there still exists two open issues: (1) The generated pseudo-labels are inevitably noisy without external supervision. (2) These is a performance gap between UDA models and the fully-supervised model. In this paper, we propose to investigate Active Learning (AL) that selects a small portion of unlabeled pixels (or images) to be annotated, which leads to an impressive performance gain. Specifically, we propose a novel Uncertainty-aware Boundary Discovery (UBD) strategy that selects the uncertain pixels in the boundary areas that contains rich contextual information. Technically, we firstly select the pixels with top entropy values, and then re-select the pixels that are exclusive to their neighbors. We leverage the Kullback-Leibler divergence between one pixel's softmax prediction and its neighbors' to measure its "exclusivity". Extensive experiments show that our approach outperforms previous methods with both pixel-level and image-level label acquisition protocols. |
Keywords | Domain Adaptation; Active Learning; Semantic Segmentation |
Contains Sensitive Content | Does not contain sensitive content |
ANZSRC Field of Research 2020 | 4602. Artificial intelligence |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Byline Affiliations | University of Electronic Science and Technology of China, China |
University of Queensland | |
Shandong Normal University, China |
https://research.usq.edu.au/item/zyx3v/pixel-exclusion-uncertainty-aware-boundary-discovery-for-active-cross-domain-semantic-segmentation
10
total views0
total downloads5
views this month0
downloads this month