RETRACTED: Dynamic Gesture Recognition Algorithm Based on 3D Convolutional Neural Network (Retracted Article)

被引:55
|
作者
Liu, Yuting [1 ,2 ]
Jiang, Du [1 ,2 ]
Duan, Haojie [2 ,3 ]
Sun, Ying [1 ,2 ,3 ]
Li, Gongfa [1 ,2 ,3 ]
Tao, Bo [1 ,2 ]
Yun, Juntong [2 ,3 ]
Liu, Ying [1 ,3 ]
Chen, Baojia [4 ]
机构
[1] Wuhan Univ Sci & Technol, Key Lab Met Equipment & Control Technol, Minist Educ, Wuhan 430081, Peoples R China
[2] Wuhan Univ Sci & Technol, Res Ctr Biomimet Robot & Intelligent Measurement, Wuhan 430081, Peoples R China
[3] Wuhan Univ Sci & Technol, Hubei Key Lab Mech Transmiss & Mfg Engn, Wuhan 430081, Peoples R China
[4] China Three Gorges Univ, Hubei Key Lab Hydroelect Machinery Design & Maint, Yichang 443002, Peoples R China
基金
中国国家自然科学基金;
关键词
Attention mechanisms - Gesture recognition algorithm - Key frame extraction technologies - Network optimization - Real time performance - Recognition accuracy - Recognition efficiency - Temporal and spatial;
D O I
10.1155/2021/4828102
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Gesture recognition is one of the important ways of human-computer interaction, which is mainly detected by visual technology. The temporal and spatial features are extracted by convolution of the video containing gesture. However, compared with the convolution calculation of a single image, multiframe image of dynamic gestures has more computation, more complex feature extraction, and more network parameters, which affects the recognition efficiency and real-time performance of the model. To solve above problems, a dynamic gesture recognition model based on CBAM-C3D is proposed. Key frame extraction technology, multimodal joint training, and network optimization with BN layer are used for making the network performance better. The experiments show that the recognition accuracy of the proposed 3D convolutional neural network combined with attention mechanism reaches 72.4% on EgoGesture dataset, which is improved greatly compared with the current main dynamic gesture recognition methods, and the effectiveness of the proposed algorithm is verified.
引用
收藏
页数:12
相关论文
共 50 条