3D Random Occlusion and Multi-layer Projection for Deep Multi-camera Pedestrian Localization

被引:10
作者
Qiu, Rui [1 ,2 ]
Xu, Ming [1 ,2 ]
Yan, Yuyao [1 ]
Smith, Jeremy S. [2 ]
Yang, Xi [1 ]
机构
[1] Xian Jiaotong Liverpool Univ, Sch Adv Technol, Suzhou 215123, Peoples R China
[2] Univ Liverpool, Dept Elect Engn & Elect, Liverpool L69 3BX, England
来源
COMPUTER VISION, ECCV 2022, PT X | 2022年 / 13670卷
基金
中国国家自然科学基金;
关键词
Multi-view detection; Deep learning; Data augmentation; Perspective transformations; TRACKING; PEOPLE;
D O I
10.1007/978-3-031-20080-9_40
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Although deep-learning based methods for monocular pedestrian detection have made great progress, they are still vulnerable to heavy occlusions. Using multi-view information fusion is a potential solution but has limited applications, due to the lack of annotated training samples in existing multi-view datasets, which increases the risk of overfitting. To address this problem, a data augmentation method is proposed to randomly generate 3D cylinder occlusions, on the ground plane, which are of the average size of pedestrians and projected to multiple views, to relieve the impact of overfitting in the training. Moreover, the feature map of each view is projected to multiple parallel planes at different heights, by using homographies, which allows the CNNs to fully utilize the features across the height of each pedestrian to infer the locations of pedestrians on the ground plane. The proposed 3DROM method has a greatly improved performance in comparison with the state-of-the-art deep learning based methods for multi-view pedestrian detection.
引用
收藏
页码:695 / 710
页数:16
相关论文
共 29 条
  • [1] Sparsity Driven People Localization with a Heterogeneous Network of Cameras
    Alahi, Alexandre
    Jacques, Laurent
    Boursier, Yannick
    Vandergheynst, Pierre
    [J]. JOURNAL OF MATHEMATICAL IMAGING AND VISION, 2011, 41 (1-2) : 39 - 58
  • [2] Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection
    Baque, Pierre
    Fleuret, Francois
    Fua, Pascal
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2017, : 271 - 279
  • [3] WILDTRACK: A Multi-camera HD Dataset for Dense Unscripted Pedestrian Detection
    Chavdarova, Tatjana
    Baque, Pierre
    Bouquet, Stephane
    Maksai, Andrii
    Jose, Cijo
    Bagautdinov, Timur
    Lettry, Louis
    Fua, Pascal
    Van Gool, Luc
    Fleuret, Francois
    [J]. 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2018, : 5030 - 5039
  • [4] Deep Multi-Camera People Detection
    Chavdarova, Tatjana
    Fleuret, Francois
    [J]. 2017 16TH IEEE INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS (ICMLA), 2017, : 848 - 853
  • [5] Tracking in a Dense Crowd Using Multiple Cameras
    Eshel, Ran
    Moses, Yael
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2010, 88 (01) : 129 - 143
  • [6] Multicamera people tracking with a probabilistic occupancy map
    Fleuret, Francois
    Berclaz, Jerome
    Lengagne, Richard
    Fua, Pascal
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2008, 30 (02) : 267 - 282
  • [7] Ge WN, 2010, LECT NOTES COMPUT SC, V6315, P324
  • [8] Deep Residual Learning for Image Recognition
    He, Kaiming
    Zhang, Xiangyu
    Ren, Shaoqing
    Sun, Jian
    [J]. 2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, : 770 - 778
  • [9] Framework for Performance Evaluation of Face, Text, and Vehicle Detection and Tracking in Video: Data, Metrics, and Protocol
    Kasturi, Rangachar
    Goldgof, Dmitry
    Soundararajan, Padmanabhan
    Manohar, Vasant
    Garofolo, John
    Bowers, Rachel
    Boonstra, Matthew
    Korzhova, Valentina
    Zhang, Jing
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2009, 31 (02) : 319 - 336
  • [10] Tracking Multiple Occluding People by Localizing on Multiple Scene Planes
    Khan, Saad M.
    Shah, Mubarak
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2009, 31 (03) : 505 - 519