Learning Instance Occlusion for Panoptic Segmentation
Abstract
Panoptic segmentation requires segments of both "things" (countable object instances) and "stuff" (uncountable and amorphous regions) within a single output. A common approach involves the fusion of instance segmentation (for "things") and semantic segmentation (for "stuff") into a non-overlapping placement of segments, and resolves overlaps. However, instance ordering with detection confidence do not correlate well with natural occlusion relationship. To resolve this issue, we propose a branch that is tasked with modeling how two instance masks should overlap one another as a binary relation. Our method, named OCFusion, is lightweight but particularly effective in the instance fusion process. OCFusion is trained with the ground truth relation derived automatically from the existing dataset annotations. We obtain state-of-the-art results on COCO and show competitive results on the Cityscapes panoptic segmentation benchmark.
Cite
Text
Lazarow et al. "Learning Instance Occlusion for Panoptic Segmentation." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020. doi:10.1109/CVPR42600.2020.01073Markdown
[Lazarow et al. "Learning Instance Occlusion for Panoptic Segmentation." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020.](https://mlanthology.org/cvpr/2020/lazarow2020cvpr-learning/) doi:10.1109/CVPR42600.2020.01073BibTeX
@inproceedings{lazarow2020cvpr-learning,
title = {{Learning Instance Occlusion for Panoptic Segmentation}},
author = {Lazarow, Justin and Lee, Kwonjoon and Shi, Kunyu and Tu, Zhuowen},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year = {2020},
doi = {10.1109/CVPR42600.2020.01073},
url = {https://mlanthology.org/cvpr/2020/lazarow2020cvpr-learning/}
}