Skeleton-based action recognition has attracted great attention in human action recognition. Existing methods for skeleton-based action recognition improve performance by designing deeper networks without considering the efficiency of the model. In this paper, we propose a simple and effective light weight graph convolutional network for skeleton-based action recognition. Our model is composed of a lightweight temporal convolutional network and spatial graph convolutional network using depthwise convolution. In addition, we propose a novel graph convolution that can take the multi-scale relationship of joints with low computational complexities. On the NTU RGB+D dataset, our proposed model achieves comparable or higher performance with much fewer parameters compared with baseline method.
|Title of host publication||2021 IEEE International Conference on Consumer Electronics-Asia, ICCE-Asia 2021|
|Publisher||Institute of Electrical and Electronics Engineers Inc.|
|Publication status||Published - 2021|
|Event||2021 IEEE International Conference on Consumer Electronics-Asia, ICCE-Asia 2021 - Gangwon, Korea, Republic of|
Duration: 2021 Nov 1 → 2021 Nov 3
|Name||2021 IEEE International Conference on Consumer Electronics-Asia, ICCE-Asia 2021|
|Conference||2021 IEEE International Conference on Consumer Electronics-Asia, ICCE-Asia 2021|
|Country/Territory||Korea, Republic of|
|Period||21/11/1 → 21/11/3|
Bibliographical noteFunding Information:
This work was supported by Hanwha Techwin.
© 2021 IEEE.
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications
- Signal Processing
- Electrical and Electronic Engineering