Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition
Traditional convolution neural networks have achieved great success in human action recognition. However, it is challenging to establish effective associations between different human bone nodes to capture detailed information. In this paper, we propose a dual attention-guided multiscale dynamic agg...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-09-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/12/10/1589 |
| _version_ | 1797552668753788928 |
|---|---|
| author | Zeyuan Hu Eung-Joo Lee |
| author_facet | Zeyuan Hu Eung-Joo Lee |
| author_sort | Zeyuan Hu |
| collection | DOAJ |
| description | Traditional convolution neural networks have achieved great success in human action recognition. However, it is challenging to establish effective associations between different human bone nodes to capture detailed information. In this paper, we propose a dual attention-guided multiscale dynamic aggregate graph convolution neural network (DAG-GCN) for skeleton-based human action recognition. Our goal is to explore the best correlation and determine high-level semantic features. First, a multiscale dynamic aggregate GCN module is used to capture important semantic information and to establish dependence relationships for different bone nodes. Second, the higher level semantic feature is further refined, and the semantic relevance is emphasized through a dual attention guidance module. In addition, we exploit the relationship of joints hierarchically and the spatial temporal correlations through two modules. Experiments with the DAG-GCN method result in good performance on the NTU-60-RGB+D and NTU-120-RGB+D datasets. The accuracy is 95.76% and 90.01%, respectively, for the cross (X)-View and X-Subon the NTU60dataset. |
| first_indexed | 2024-03-10T16:04:20Z |
| format | Article |
| id | doaj.art-103a2d2e83444cb9b20fd3292dd5e18a |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-03-10T16:04:20Z |
| publishDate | 2020-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-103a2d2e83444cb9b20fd3292dd5e18a2023-11-20T15:00:10ZengMDPI AGSymmetry2073-89942020-09-011210158910.3390/sym12101589Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action RecognitionZeyuan Hu0Eung-Joo Lee1Department of Information Communication Engineering, Tongmyong University, Busan 48520, KoreaDepartment of Information Communication Engineering, Tongmyong University, Busan 48520, KoreaTraditional convolution neural networks have achieved great success in human action recognition. However, it is challenging to establish effective associations between different human bone nodes to capture detailed information. In this paper, we propose a dual attention-guided multiscale dynamic aggregate graph convolution neural network (DAG-GCN) for skeleton-based human action recognition. Our goal is to explore the best correlation and determine high-level semantic features. First, a multiscale dynamic aggregate GCN module is used to capture important semantic information and to establish dependence relationships for different bone nodes. Second, the higher level semantic feature is further refined, and the semantic relevance is emphasized through a dual attention guidance module. In addition, we exploit the relationship of joints hierarchically and the spatial temporal correlations through two modules. Experiments with the DAG-GCN method result in good performance on the NTU-60-RGB+D and NTU-120-RGB+D datasets. The accuracy is 95.76% and 90.01%, respectively, for the cross (X)-View and X-Subon the NTU60dataset.https://www.mdpi.com/2073-8994/12/10/1589human action recognitionmultiscale graph convolutional networksdynamic aggregationhierarchical level semantic informationspatial and temporal correlation |
| spellingShingle | Zeyuan Hu Eung-Joo Lee Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition Symmetry human action recognition multiscale graph convolutional networks dynamic aggregation hierarchical level semantic information spatial and temporal correlation |
| title | Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition |
| title_full | Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition |
| title_fullStr | Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition |
| title_full_unstemmed | Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition |
| title_short | Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition |
| title_sort | dual attention guided multiscale dynamic aggregate graph convolutional networks for skeleton based human action recognition |
| topic | human action recognition multiscale graph convolutional networks dynamic aggregation hierarchical level semantic information spatial and temporal correlation |
| url | https://www.mdpi.com/2073-8994/12/10/1589 |
| work_keys_str_mv | AT zeyuanhu dualattentionguidedmultiscaledynamicaggregategraphconvolutionalnetworksforskeletonbasedhumanactionrecognition AT eungjoolee dualattentionguidedmultiscaledynamicaggregategraphconvolutionalnetworksforskeletonbasedhumanactionrecognition |