Weakly Supervised Segmentation of Small Buildings with Point Labels
Abstract
Most supervised image segmentation methods require delicate and time-consuming pixel-level labeling of building or objects, especially for small objects. In this paper, we present a weakly supervised segmentation network for aerial/satellite images, separately considering small and large objects. First, we propose a simple point labeling method for small objects, while large objects are fully labeled. Then, we present a segmentation network trained with a small object mask to separate small and large objects in the loss function. During training, we employ a memory bank to cope with the limited number of point labels. Experiments results with three public datasets demonstrate the feasibility of our approach.
Cite
Text
Lee et al. "Weakly Supervised Segmentation of Small Buildings with Point Labels." International Conference on Computer Vision, 2021. doi:10.1109/ICCV48922.2021.00731Markdown
[Lee et al. "Weakly Supervised Segmentation of Small Buildings with Point Labels." International Conference on Computer Vision, 2021.](https://mlanthology.org/iccv/2021/lee2021iccv-weakly/) doi:10.1109/ICCV48922.2021.00731BibTeX
@inproceedings{lee2021iccv-weakly,
title = {{Weakly Supervised Segmentation of Small Buildings with Point Labels}},
author = {Lee, Jae-Hun and Kim, ChanYoung and Sull, Sanghoon},
booktitle = {International Conference on Computer Vision},
year = {2021},
pages = {7406-7415},
doi = {10.1109/ICCV48922.2021.00731},
url = {https://mlanthology.org/iccv/2021/lee2021iccv-weakly/}
}