MPE-HRNetL: A Lightweight High-Resolution Network for Multispecies Animal Pose Estimation

被引:0
|
作者
Shen, Jiquan [1 ,2 ]
Jiang, Yaning [3 ]
Luo, Junwei [1 ]
Wang, Wei [4 ]
机构
[1] Henan Polytech Univ, Sch Software, Jiaozuo 454000, Peoples R China
[2] Anyang Inst Technol, Anyang 455000, Peoples R China
[3] Henan Polytech Univ, Sch Comp Sci & Technol, Jiaozuo 454000, Peoples R China
[4] CSSC Haifeng Aviat Technol Co Ltd, 4,Xinghuo Rd,Fengtai Sci Pk, Beijing 100070, Peoples R China
基金
中国国家自然科学基金;
关键词
animal pose estimation; multiscale mixed attention mechanism; high-resolution network; spatial pyramid pooling; Lite-HRNet; BEHAVIOR;
D O I
10.3390/s24216882
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
Animal pose estimation is crucial for animal health assessment, species protection, and behavior analysis. It is an inevitable and unstoppable trend to apply deep learning to animal pose estimation. In many practical application scenarios, pose estimation models must be deployed on edge devices with limited resource. Therefore, it is essential to strike a balance between model complexity and accuracy. To address this issue, we propose a lightweight network model, i.e., MPE-HRNet.L, by improving Lite-HRNet. The improvements are threefold. Firstly, we improve Spatial Pyramid Pooling-Fast and apply it and the improved version to different branches. Secondly, we construct a feature extraction module based on a mixed pooling module and a dual spatial and channel attention mechanism, and take the feature extraction module as the basic module of MPE-HRNet.L. Thirdly, we introduce a feature enhancement stage to enhance important features. The experimental results on the AP-10K dataset and the Animal Pose dataset verify the effectiveness and efficiency of MPE-HRNet.L.
引用
收藏
页数:16
相关论文
共 50 条
  • [21] High-resolution network with an auxiliary channel for 2D hand pose estimation
    Pan, Tianhong
    Wang, Zheng
    MULTIMEDIA TOOLS AND APPLICATIONS, 2023, 83 (12) : 36683 - 36694
  • [22] DB-HRNet: Dual Branch High-Resolution Network for Human Pose Estimation
    Wang, Yanxia
    Wang, Renjie
    Shi, Hu
    IEEE ACCESS, 2023, 11 : 120628 - 120641
  • [23] An Improved High-Resolution Network-Based Method for Yoga-Pose Estimation
    Li, Jianrong
    Zhang, Dandan
    Shi, Lei
    Ke, Ting
    Zhang, Chuanlei
    APPLIED SCIENCES-BASEL, 2023, 13 (15):
  • [24] Research on Multi-Target Animal Pose Estimation Based on Improved High Resolution Network
    Xu, Guidong
    Xu, Yang
    Deng, Hui
    Mo, Han
    Computer Engineering and Applications, 2023, 59 (22) : 182 - 192
  • [25] Lite CSW-HRNet: Lightweight High-Resolution Human Pose Estimation Based on Channel Spatial Weighting
    Xi, Yang
    Zhang, Zi-Hao
    Meng, Si-Yu
    Fu, Jia
    Wu, Zhen-Yu
    Wang, Wen-Jing
    Journal of Network Intelligence, 2024, 9 (03): : 1641 - 1656
  • [26] Lite-HRNet: A Lightweight High-Resolution Network
    Yu, Changqian
    Xiao, Bin
    Gao, Changxin
    Yuan, Lu
    Zhang, Lei
    Sang, Nong
    Wang, Jingdong
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 10435 - 10445
  • [27] MS-HRNet: multi-scale high-resolution network for human pose estimation
    Wang, Yanxia
    Wang, Renjie
    Shi, Hu
    Liu, Dan
    JOURNAL OF SUPERCOMPUTING, 2024, 80 (12): : 17269 - 17291
  • [28] Deep High-Resolution Representation Learning for Human Pose Estimation
    Sun, Ke
    Xiao, Bin
    Liu, Dong
    Wang, Jingdong
    2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 5686 - 5696
  • [29] SHaRPose: Sparse High-Resolution Representation for Human Pose Estimation
    An, Xiaoqi
    Zhao, Lin
    Gong, Chen
    Wang, Nannan
    Wang, Di
    Yang, Jian
    THIRTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 38 NO 2, 2024, : 691 - 699
  • [30] HR-xNet: A Novel High-Resolution Network for Human Pose Estimation with Low Resource Consumption
    Feng, Cun
    Zhang, Rong
    Guo, Lijun
    2024 IEEE 18TH INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION, FG 2024, 2024,