Multi-label convolutional neural network based pedestrian attribute classification

doi:10.1016/j.imavis.2016.07.004

CASIA OpenIR > 模式识别国家重点实验室 > 生物识别与安全技术研究

	Multi-label convolutional neural network based pedestrian attribute classification
	Zhu, Jianqing 1; Liao, Shengcai2 ; Lei, Zhen2 ; Li, Stan Z.2
发表期刊	IMAGE AND VISION COMPUTING
	2017-02-01
卷号	58 页码:224-229
文章类型	Article
摘要	Recently, pedestrian attributes like gender, age, clothing etc., have been used as soft biometric traits for recognizing people. Unlike existing methods that assume the independence of attributes during their prediction, we propose a multi-label convolutional neural network(MLCNN) to predict multiple attributes together in a unified framework. Firstly, a pedestrian image is roughly divided into multiple overlapping body parts, which are simultaneously integrated in the multi-label convolutional neural network. Secondly, these parts are filtered independently and aggregated in the cost layer. The cost function is a combination of multiple binary attribute classification cost functions. Experiments show that the proposed method significantly outperforms the SVM based method on the PETA database. (C) 2016 Elsevier B.V. All rights reserved.
关键词	Pedestrian Attribute Classification Multi-label Classification Convolutional Neural Network
WOS标题词	Science & Technology ; Technology ; Physical Sciences
DOI	10.1016/j.imavis.2016.07.004
关键词[WOS]	SOFT BIOMETRICS
收录类别	SCI
语种	英语
项目资助者	Scientific Research Funds of Huaqiao University(16BS108) ; National Natural Science Foundation of China(61602191 ; Chinese Academy of Sciences Project(KGZD-EW-102-2) ; 61375037 ; 61473291 ; 61572501 ; 61502491 ; 61572536)
WOS研究方向	Computer Science ; Engineering ; Optics
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Software Engineering ; Computer Science, Theory & Methods ; Engineering, Electrical & Electronic ; Optics
WOS记录号	WOS:000395844700021
引用统计	被引频次：7[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/14433
专题	模式识别国家重点实验室_生物识别与安全技术研究
作者单位	1.Huaqiao Univ, Coll Engn, Quanzhou 362021, Fujian, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Ctr Biometr & Secur Res, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhu, Jianqing,Liao, Shengcai,Lei, Zhen,et al. Multi-label convolutional neural network based pedestrian attribute classification[J]. IMAGE AND VISION COMPUTING,2017,58:224-229.
APA	Zhu, Jianqing,Liao, Shengcai,Lei, Zhen,&Li, Stan Z..(2017).Multi-label convolutional neural network based pedestrian attribute classification.IMAGE AND VISION COMPUTING,58,224-229.
MLA	Zhu, Jianqing,et al."Multi-label convolutional neural network based pedestrian attribute classification".IMAGE AND VISION COMPUTING 58(2017):224-229.

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
JQZHU-IVC-2017.pdf（1036KB）	期刊论文	作者接受稿	开放获取	CC BY-NC-SA	浏览下载