Weakly Supervised Minirhizotron Image Segmentation with MIL-CAM
Abstract
We present a multiple instance learning class activation map (MIL-CAM) approach for pixel-level minirhizotron image segmentation given weak image-level labels. Minirhizotrons are used to image plant roots in situ. Minirhizotron imagery is often composed of soil containing a few long and thin root objects of small diameter. The roots prove to be challenging for existing semantic image segmentation methods to discriminate. In addition to learning from weak labels, our proposed MIL-CAM approach re-weights the root versus soil pixels during analysis for improved performance due to the heavy imbalance between soil and root pixels. The proposed approach outperforms other attention map and multiple instance learning methods for localization of root objects in minirhizotron imagery.
Cite
Text
Yu et al. "Weakly Supervised Minirhizotron Image Segmentation with MIL-CAM." European Conference on Computer Vision Workshops, 2020. doi:10.1007/978-3-030-65414-6_30Markdown
[Yu et al. "Weakly Supervised Minirhizotron Image Segmentation with MIL-CAM." European Conference on Computer Vision Workshops, 2020.](https://mlanthology.org/eccvw/2020/yu2020eccvw-weakly/) doi:10.1007/978-3-030-65414-6_30BibTeX
@inproceedings{yu2020eccvw-weakly,
title = {{Weakly Supervised Minirhizotron Image Segmentation with MIL-CAM}},
author = {Yu, Guohao and Zare, Alina and Xu, Weihuang and Matamala, Roser and Reyes-Cabrera, Joel and Fritschi, Felix B. and Juenger, Thomas E.},
booktitle = {European Conference on Computer Vision Workshops},
year = {2020},
pages = {433-449},
doi = {10.1007/978-3-030-65414-6_30},
url = {https://mlanthology.org/eccvw/2020/yu2020eccvw-weakly/}
}