Adaptive Multi-Scale Difference Graph Convolution Network for Skeleton-Based Action Recognition

被引:2
|
作者
Wang, Xiaojuan [1 ]
Gan, Ziliang [1 ]
Jin, Lei [1 ]
Xiao, Yabo [1 ]
He, Mingshu [1 ]
机构
[1] Beijing Univ Posts & Telecommun, Sch Elect Engn, 10 Xitucheng Rd, Beijing 100876, Peoples R China
基金
中国国家自然科学基金;
关键词
skeleton-based action recognition; graph convolution networks; difference convolution;
D O I
10.3390/electronics12132852
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Graph convolutional networks (GCNs) have obtained remarkable performance in skeleton-based action recognition. However, previous approaches fail to capture the implicit correlations between joints and handle actions across varying time intervals. To address these problems, we propose an adaptive multi-scale difference graph convolution Network (AMD-GCN), which comprises an adaptive spatial graph convolution module (ASGC) and a multi-scale temporal difference convolution module (MTDC). The first module is capable of acquiring data-dependent and channel-wise graphs that are adaptable to both samples and channels. The second module employs the multi-scale approach to model temporal information across a range of time scales. Additionally, the MTDC incorporates an attention-enhanced module and difference convolution to accentuate significant channels and enhance temporal features, respectively. Finally, we propose a multi-stream framework for integrating diverse skeletal modalities to achieve superior performance. Our AMD-GCN approach was extensively tested and proven to outperform the current state-of-the-art methods on three widely recognized benchmarks: the NTU-RGB+D, NTU-RGB+D 120, and Kinetics Skeleton datasets.
引用
收藏
页数:19
相关论文
共 50 条
  • [31] Multi-stream adaptive 3D attention graph convolution network for skeleton-based action recognition
    Lubin Yu
    Lianfang Tian
    Qiliang Du
    Jameel Ahmed Bhutto
    Applied Intelligence, 2023, 53 : 14838 - 14854
  • [32] Multi-Part Adaptive Graph Convolutional Network for Skeleton-Based Action Recognition
    Wang, Wei
    Xie, Wei
    Tu, Zhigang
    Li, Wanxin
    Jin, Lianghao
    2022 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2022,
  • [33] Self-Relational Graph Convolution Network for Skeleton-Based Action Recognition
    Yussif, Sophyani Banaamwini
    Xie, Ning
    Yang, Yang
    Shen, Heng Tao
    PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2023, 2023, : 27 - 36
  • [34] Mixed graph convolution and residual transformation network for skeleton-based action recognition
    Shuhua Liu
    Xiaoying Bai
    Ming Fang
    Lanting Li
    Chih-Cheng Hung
    Applied Intelligence, 2022, 52 : 1544 - 1555
  • [35] Temporal-enhanced graph convolution network for skeleton-based action recognition
    Xie, Yulai
    Zhang, Yang
    Ren, Fang
    IET COMPUTER VISION, 2022, 16 (03) : 266 - 279
  • [36] Mixed graph convolution and residual transformation network for skeleton-based action recognition
    Liu, Shuhua
    Bai, Xiaoying
    Fang, Ming
    Li, Lanting
    Hung, Chih-Cheng
    APPLIED INTELLIGENCE, 2022, 52 (02) : 1544 - 1555
  • [37] Kernel Attention Based Multi-scale Adaptive Graph Convolutional Neural Network for Skeleton-Based
    Liu, Yanan
    Zhang, Hao
    Xu, Dan
    2021 IEEE 7TH INTERNATIONAL CONFERENCE ON VIRTUAL REALITY (ICVR 2021), 2021, : 96 - 103
  • [38] Spatial adaptive graph convolutional network for skeleton-based action recognition
    Zhu, Qilin
    Deng, Hongmin
    APPLIED INTELLIGENCE, 2023, 53 (14) : 17796 - 17808
  • [39] Glimpse and focus: Global and local-scale graph convolution network for skeleton-based action recognition
    Gao, Xuehao
    Du, Shaoyi
    Yang, Yang
    NEURAL NETWORKS, 2023, 167 : 551 - 558
  • [40] Spatial adaptive graph convolutional network for skeleton-based action recognition
    Qilin Zhu
    Hongmin Deng
    Applied Intelligence, 2023, 53 : 17796 - 17808