计算机应用与软件2024，Vol.41Issue(7)：165-170,254,7.DOI:10.3969/j.issn.1000-386x.2024.07.025
基于时空注意图卷积的人体动作识别
ACTION RECOGNITION BASED ON SPATIAL-TEMPORAL ATTENTION GRAPH CONVOLUTION NEURAL NETWORK
摘要
Abstract
In view of the low application of key joints and features in human action recognition based on skeleton data,an improved action recognition system based on the fusion of spatial-temporal graph convolution neural network and channel-spatial union attention block is proposed.The structural features were obtained by spatial graph convolution,and the key joints and key structure information were enhanced by channel-spatial union attention module.The advanced spatial-temporal features were obtained by time graph convolution.The recognition results were obtained by global pooling layer and Softmax classifier.The experimental results show that while the key joints and structural features are enhanced,the original feature information is retained.This algorithm has higher accuracy in skeleton-based action recog-nition.关键词
人体动作识别/骨骼数据/注意力模块/关键节点/时空图卷积Key words
Human action recognition/Skeleton data/Attention module/Key joints/Spatial-temporal graph convo-lution neural network分类
信息技术与安全科学引用本文复制引用
赵登阁,智敏..基于时空注意图卷积的人体动作识别[J].计算机应用与软件,2024,41(7):165-170,254,7.基金项目
内蒙古自治区高等学校科学研究项目(NJZ21004) （NJZ21004）
内蒙古自然科学基金项目(2018MS06008). （2018MS06008）
