Mirrored Non-Maximum Suppression for Accurate Object Part Localization

CASIA OpenIR > 智能感知与计算研究中心

	Mirrored Non-Maximum Suppression for Accurate Object Part Localization
	Fu LR(付连锐); Junge Zhang; Kaiqi Huang
	2015-11
会议名称	3rd IAPR Asian Conference on Pattern Recognition
会议录名称	Proceeding of 3rd IAPR Asian Conference on Pattern Recognition
页码	51-55
会议日期	2015.11.03-2015.11.06
会议地点	Kuala Lumpur, Malaysia
摘要	There has been significant progress in object part localization such as human pose estimation and facial landmark detection. In most of the previous methods, two phenomena are ignored. Firstly, they usually output a set of candidate pose hypotheses but the hypothesis with the highest score obtained by Non-Maximum Suppression (NMS) is not always the optimal result. Secondly, they can not get exactly bilaterally symmetric keypoints on the mirrored images even though the training data is always augmented with mirrored images. In fact, the intrinsic relationship between the original image and the mirrored one is helpful for object part localization. In this paper, we propose Mirrored Non-Maximum Suppression (Mirrored NMS) which can utilize mirrored detections to improve the accuracy of object part localization. Experimental results show that our method can improve the state-of-the-art accuracy by 1.3∼3.0% in PCP for human pose estimation and can produce more accurate results than averaging multiple hypotheses for facial landmark detection.
关键词	Non-maximum
收录类别	EI
语种	英语
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/11650
专题	智能感知与计算研究中心
通讯作者	Kaiqi Huang
作者单位	中国科学院自动化研究所
第一作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Fu LR,Junge Zhang,Kaiqi Huang. Mirrored Non-Maximum Suppression for Accurate Object Part Localization[C],2015:51-55.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
[ACPR15_Fu]Mirrored （3090KB）	会议论文		开放获取	CC BY-NC-SA	浏览下载