Rethinking Segmentation Guidance for Weakly Supervised Object Detection

Ke Yang, Peng Zhang, Peng Qiao, Zhiyuan Wang, Huadong Dai, Tianlong Shen, Dongsheng Li, Yong Dou; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2020, pp. 946-947

Weakly supervised object detection aims at learning object detectors with only image-level category labels. Most existing methods tend to solve this problem by using a multiple instance learning detector which is usually trapped to discriminate object parts, rather than the entire object. In order to select high-quality proposals, recent works leverage objectness scores derived from weakly-supervised segmentation maps to rank the object proposals. Base our observation, this kind of segmentation guided method always fails due to neglect of the fact that objectness of all proposals inside the ground-truth box should be consistent. In this paper, we propose a novel object representation named Objectness Consistent Representation (OCR) to meet the consistency criterion of objectness. Specifically, we project the segmentation confidence scores into two orthogonal directions, namely vertical and horizontal, to get the OCR. With the novel object representation, more high-quality proposals can be mined for learning a much stronger object detector.