HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis

被引:399
作者
Liu, Xihui [1 ,2 ]
Zhao, Haiyu [2 ]
Tian, Maoqing [2 ]
Sheng, Lu [1 ]
Shao, Jing [2 ]
Yi, Shuai [2 ]
Yan, Junjie [2 ]
Wang, Xiaogang [1 ]
机构
[1] Chinese Univ Hong Kong, Sha Tin, Hong Kong, Peoples R China
[2] SenseTime Grp Ltd, Hong Kong, Hong Kong, Peoples R China
来源
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV) | 2017年
关键词
D O I
10.1109/ICCV.2017.46
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Pedestrian analysis plays a vital role in intelligent video surveillance and is a key component for security-centric computer vision systems. Despite that the convolutional neural networks are remarkable in learning discriminative features from images, the learning of comprehensive features of pedestrians for fine-grained tasks remains an open problem. In this study, we propose a new attention-based deep neural network, named as HydraPlus-Net (HPnet), that multi-directionally feeds the multi-level attention maps to different feature layers. The attentive deep features learned from the proposed HP-net bring unique advantages: (1) the model is capable of capturing multiple attentions from low-level to semantic-level, and (2) it explores the multi-scale selectiveness of attentive features to enrich the final feature representations for a pedestrian image. We demonstrate the effectiveness and generality of the proposed HP-net for pedestrian analysis on two tasks, i.e. pedestrian attribute recognition and person re-identification. Intensive experimental results have been provided to prove that the HP-net outperforms the state-of-the-art methods on various datasets.(1)
引用
收藏
页码:350 / 359
页数:10
相关论文
共 37 条
[1]  
[Anonymous], ARXIV160300370
[2]  
[Anonymous], 2017, CVPR
[3]  
[Anonymous], 2014, arXiv
[4]  
[Anonymous], 2015, ARXIV151205300
[5]  
[Anonymous], 2016, CVPR
[6]  
[Anonymous], P 22 ACM INT C MULT
[7]  
[Anonymous], 2010, BMVC
[8]  
[Anonymous], 2015, CVPR
[9]  
[Anonymous], 2016, ECCV
[10]  
[Anonymous], 2017, CVPR