Using Segmentation to Predict the Absence of Occluded Parts
Golnaz Ghiasi and Charless Fowlkes
Abstract
Occlusion poses a significant difficulty for detecting and localizing object keypoints and subsequent fine-grained identification. We propose a part-based face detection model that utilizes bottom-up class-specific segmentation in order to jointly detect and segment out the foreground pixels belonging to the face. The model explicitly represents occlusion of parts at the detection phase, allowing for hypothesized figure-ground segmentation to suggest coherent patterns of part occlusion. We show that this bi-directional interaction between recognition and grouping results in state-of-the-art part localization accuracy for challenging benchmarks with significant occlusion and yields substantial gains in the precision of keypoint occlusion prediction.
Session
Poster 1
Files
Extended Abstract (PDF, 1041K)
Paper (PDF, 6M)
DOI
10.5244/C.29.22
https://dx.doi.org/10.5244/C.29.22
Citation
Golnaz Ghiasi and Charless Fowlkes. Using Segmentation to Predict the Absence of Occluded Parts. In Xianghua Xie, Mark W. Jones, and Gary K. L. Tam, editors, Proceedings of the British Machine Vision Conference (BMVC), pages 22.1-22.12. BMVA Press, September 2015.
Bibtex
@inproceedings{BMVC2015_22,
title={Using Segmentation to Predict the Absence of Occluded Parts},
author={Golnaz Ghiasi and Charless Fowlkes},
year={2015},
month={September},
pages={22.1-22.12},
articleno={22},
numpages={12},
booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
publisher={BMVA Press},
editor={Xianghua Xie, Mark W. Jones, and Gary K. L. Tam},
doi={10.5244/C.29.22},
isbn={1-901725-53-7},
url={https://dx.doi.org/10.5244/C.29.22}
}