Using segmentation to predict the absence of occluded parts
icon 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.

Download: pdf

Text Reference

Golnaz Ghiasi and Charless C. Fowlkes. Using segmentation to predict the absence of occluded parts. In British Machine Vision Conference (BMVC). 2015.

BibTeX Reference

@inproceedings{GhiasiF_BMVC_2015,
    author = "Ghiasi, Golnaz and Fowlkes, Charless C.",
    title = "Using segmentation to predict the absence of occluded parts",
    booktitle = "British Machine Vision Conference (BMVC)",
    year = "2015",
    tag = "object_recognition,people,grouping"
}