SSR-HEF: Crowd Counting With Multiscale Semantic Refining and Hard Example Focusing

doi:10.1109/TII.2022.3160634

	SSR-HEF: Crowd Counting With Multiscale Semantic Refining and Hard Example Focusing
	Chen, Jiwei 1,2,3; Wang, Kewei 4; Su, Wen 5; Wang, Zengfu 1,2
刊名	IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
	2022-10-01
卷号	18
关键词	Semantics Task analysis Feature extraction Focusing Informatics Estimation Prediction algorithms Crowd counting density map hard example focusing (HEF) multiscale semantic refining strategy (SSR)
ISSN号	1551-3203
DOI	10.1109/TII.2022.3160634
通讯作者	Wang, Zengfu(zfwang@ustc.edu.cn)
英文摘要	Crowd counting based on density maps is generally regarded as a regression task. Deep learning is used to learn the mapping between image content and crowd density distribution. Although great success has been achieved, some pedestrians far away from the camera are difficult to be detected. And the number of hard examples is often larger. Existing methods with simple Euclidean distance algorithm indiscriminately optimize the hard and easy examples so that the densities of hard examples are usually incorrectly predicted to be lower or even zero, which results in large counting errors. To address this problem, we are the first to propose the hard example focusing (HEF) algorithm for the regression task of crowd counting. The HEF algorithm makes our model rapidly focus on hard examples by attenuating the contribution of easy examples. Then higher importance will be given to the hard examples with wrong estimations. Moreover, the scale variations in crowd scenes are large, and the scale annotations are labor-intensive and expensive. By proposing a multiscale semantic refining strategy, lower layers of our model can break through the limitation of deep learning to capture semantic features of different scales to sufficiently deal with the scale variation. We perform extensive experiments on six benchmark datasets to verify the proposed method. Results indicate the superiority of our proposed method over the state-of-the-art methods. Moreover, our designed model is smaller and faster.
资助项目	Special Project of Strategic Leading Science and Technology of Chinese Academy of Sciences[XDC08020000] ; Special Project of Strategic Leading Science and Technology of Chinese Academy of Sciences[XDC08020400] ; National Natural Science Foundation of China[61472393] ; National Natural Science Foundation of China[TII-21-1912]
WOS关键词	PEOPLE
WOS研究方向	Automation & Control Systems ; Computer Science ; Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000838389400008
资助机构	Special Project of Strategic Leading Science and Technology of Chinese Academy of Sciences ; National Natural Science Foundation of China
内容类型	期刊论文
源URL	[http://ir.hfcas.ac.cn:8080/handle/334002/131808]
专题	中国科学院合肥物质科学研究院
通讯作者	Wang, Zengfu
作者单位	1.Chinese Acad Sci, Inst Intelligent Machines, Hefei Inst Phys Sci, Hefei 230031, Peoples R China 2.Univ Sci & Technol China, Hefei 230026, Peoples R China 3.Hefei Univ Technol, Hefei 230009, Peoples R China 4.Univ Sydney, Fac Engn, Sch Comp Sci, Camperdown, NSW 2006, Australia 5.Zhejiang Sci Tech Univ, Virtual Real Lab, Hangzhou 314423, Peoples R China
推荐引用方式 GB/T 7714	Chen, Jiwei,Wang, Kewei,Su, Wen,et al. SSR-HEF: Crowd Counting With Multiscale Semantic Refining and Hard Example Focusing[J]. IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS,2022,18.
APA	Chen, Jiwei,Wang, Kewei,Su, Wen,&Wang, Zengfu.(2022).SSR-HEF: Crowd Counting With Multiscale Semantic Refining and Hard Example Focusing.IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS,18.
MLA	Chen, Jiwei,et al."SSR-HEF: Crowd Counting With Multiscale Semantic Refining and Hard Example Focusing".IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 18(2022).