Adaptive Binarization for Weakly Supervised Affordance Segmentation
Abstract
The concept of affordance is important to understand the relevance of object parts for a certain functional interaction. Affordance types generalize across object categories and are not mutually exclusive. This makes the segmentation of affordance regions of objects in images a difficult task. In this work, we build on an iterative approach that learns a convolutional neural network for affordance segmentation from sparse keypoints. During this process, the predictions of the network need to be binarized. To this end, we propose an adaptive approach for binarization and estimate the parameters for initialization by approximated cross validation. We evaluate our approach on two affor-dance datasets where our approach outperforms the state-of-the-art for weakly supervised affordance segmentation.
Cite
Text
Gall and Sawatzky. "Adaptive Binarization for Weakly Supervised Affordance Segmentation." IEEE/CVF International Conference on Computer Vision Workshops, 2017. doi:10.1109/ICCVW.2017.164Markdown
[Gall and Sawatzky. "Adaptive Binarization for Weakly Supervised Affordance Segmentation." IEEE/CVF International Conference on Computer Vision Workshops, 2017.](https://mlanthology.org/iccvw/2017/gall2017iccvw-adaptive/) doi:10.1109/ICCVW.2017.164BibTeX
@inproceedings{gall2017iccvw-adaptive,
title = {{Adaptive Binarization for Weakly Supervised Affordance Segmentation}},
author = {Gall, Juergen and Sawatzky, Johann},
booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
year = {2017},
pages = {1383-1391},
doi = {10.1109/ICCVW.2017.164},
url = {https://mlanthology.org/iccvw/2017/gall2017iccvw-adaptive/}
}