摘要
随着深度学习技术的发展,肌电动作识别技术取得许多显著的成就。然而,单一模态的肌电数据难以充分描述肌电信号中所蕴含的运动意图。该研究围绕肌电手势识别技术,提出基于多模态数据融合的动作识别方法,对方法的识别正确率及时间消耗进行重点分析,并与基于单一模态的动作识别方法进行比较,以验证所提出方法的有效性。该研究为肌电动作识别技术的发展提供有益的理论参考。
With the development of deep learning technology,electromyographic(EMG)action recognition technology has achieved many significant achievements.However,single mode EMG data is difficult to fully describe the movement intention contained in EMG signals.This research focuses on the EMG gesture recognition technology,proposes a motion recognition method based on multi-modal data fusion,focuses on the analysis of the recognition accuracy and time consumption of the method,and compares it with the action recognition method based on single mode to verify the effectiveness of the proposed method.This study will provide useful theoretical references for the development of electromyographic action recognition technology.
作者
张安元
周超然
高飞
王宇
ZHANG Anyuan;ZHOU Chaoran;GAO Fei
出处
《科技创新与应用》
2025年第23期38-40,44,共4页
Technology Innovation and Application
基金
2024年吉林省高教科研课题(JGJX24D0118)。
关键词
肌电
手势识别
多模态
深度学习
运动姿态
electromyographic(EMG)
gesture recognition
multimodal
Deep Learning
exercise posture
作者简介
第一作者:张安元(1992-),男,博士,讲师。研究方向为人机交互。
