Self-attention Guidance Based Crowd Localization and Counting

被引:1
作者
Ma, Zhouzhou [1 ,2 ]
Gu, Guanghua [1 ,2 ]
Zhao, Wenrui [1 ,2 ]
机构
[1] Yanshan Univ, Sch Informat Sci & Engn, Qinhuangdao 066000, Peoples R China
[2] Hebei Key Lab Informat Transmiss & Signal Proc, Qinhuangdao 066000, Peoples R China
基金
中国国家自然科学基金;
关键词
Crowd localization; crowd counting; transformer; point supervision; object detection; IMAGE; NETWORK;
D O I
10.1007/s11633-023-1428-6
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Most existing studies on crowd analysis are limited to the level of counting, which cannot provide the exact location of individuals. This paper proposes a self-attention guidance based crowd localization and counting network (SA-CLCN), which can simultaneously locate and count crowds. We take the form of object detection, using the original point annotations of crowd datasets as supervision to train the network. Ultimately, the center point coordinate of each head as well as the number of crowds are predicted. Specifically, to cope with the spatial and positional variations of the crowd, the proposed method introduces transformer to construct a globallocal feature extractor (GLFE) together with the convolutional structure. It establishes the near-to-far dependency between elements so that the global context and local detail features of the crowd image can be extracted simultaneously. Then, this paper designs a pyramid feature fusion module (PFFM) to fuse the global and local information from high level to low level to obtain a multiscale feature representation. In downstream tasks, this paper predicts candidate point offsets and confidence scores by a simple regression header and classification header. In addition, the Hungarian algorithm is used to match the predicted point set and the labelled point set to facilitate the calculation of losses. The proposed network avoids the errors or higher costs associated with using traditional density maps or bounding box annotations. Importantly, we have conducted extensive experiments on several crowd datasets, and the proposed method has produced competitive results in both counting and localization.
引用
收藏
页码:966 / 982
页数:17
相关论文
共 50 条
  • [31] Crowd counting method based on cross column fusion attention mechanism
    Cui, Xiao
    Zhang, Zhi-Feng
    Zheng, Qian
    Cao, Jie
    JOURNAL OF ELECTRONIC IMAGING, 2021, 30 (03)
  • [32] Crowd counting in complex scenes based on an attention aware CNN network
    Li, Zhaoxin
    Lu, Shuhua
    Lan, Lingqiang
    Liu, Qiyuan
    JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION, 2022, 87
  • [33] ACCNet: Attention-based Contextual Convolutional Network for Crowd Counting
    Huang, Yaoying
    Zhu, Aichun
    Duan, Guoxiu
    Hu, Fangqiang
    Li, Yifeng
    2020 CHINESE AUTOMATION CONGRESS (CAC 2020), 2020, : 1926 - 1931
  • [34] Context Attention Fusion Network for crowd counting
    Wang, Tao
    Zhang, Ting
    Zhang, Kaibing
    Wang, Huake
    Li, Minqi
    Lu, Jian
    KNOWLEDGE-BASED SYSTEMS, 2023, 271
  • [35] ATTENTION GUIDED REGION DIVISION FOR CROWD COUNTING
    Pan, Xiaoqi
    Mo, Hong
    Zhou, Zhong
    Wu, Wei
    2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2020, : 2568 - 2572
  • [36] TRIPLE ATTENTION FOR ROBUST VIDEO CROWD COUNTING
    Wu, Qiyao
    Zhang, Chongyang
    Kong, Xiyu
    Zhao, Muming
    Chen, Yanjun
    2020 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2020, : 1966 - 1970
  • [37] DENSE POINT PREDICTION: A SIMPLE BASELINE FOR CROWD COUNTING AND LOCALIZATION
    Wang, Yi
    Hou, Xinyu
    Chou, Lap-Pui
    2021 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA & EXPO WORKSHOPS (ICMEW), 2021,
  • [38] Crowd Counting and Individual Localization Using Pseudo Square Label
    Ryu, Jihye
    Song, Kwangho
    IEEE ACCESS, 2024, 12 : 68160 - 68170
  • [39] Local Point Matching Network for Stabilized Crowd Counting and Localization
    Niu, Lin
    Wang, Xinggang
    Duan, Chen
    Shen, Qiongxia
    Liu, Wenyu
    PATTERN RECOGNITION AND COMPUTER VISION, PT I, PRCV 2022, 2022, 13534 : 566 - 579
  • [40] WEAKLY SUPERVISED CROWD-WISE ATTENTION FOR ROBUST CROWD COUNTING
    Kong, Xiyu
    Zhao, Muming
    Zhou, Hao
    Zhang, Chongyang
    2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2020, : 2722 - 2726