Convolutional neural network and subdomain adaptive low-sampling-rate myoelectricity gesture recognition method

• Main Authors: ZHOU DIAO, ZHANG QI, ZONG JING, ZHOU JIANHUA
• Format: Patent
• Language: Chinese; English
• Published: 29.12.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention relates to the technical field of gesture recognition, and discloses a low-sampling-rate myoelectric gesture recognition method adaptive to a convolutional neural network and a subdomain, and the method comprises a data collection module, a data preprocessing module, a gesture recognition module and an evaluation analysis module, the gesture recognition module is divided into three stages, and a data input stage, an information expansion stage and a model prediction stage, the data input stage is used for connecting the data output by the data preprocessing module, and the information expansion stage is used for expanding the data in different branches. According to the low-sampling-rate myoelectric gesture recognition method adaptive by combining the convolutional neural network with the sub-domain, verification is carried out on two databases DB1 and DB5 in NinaPro, and the result shows that in low-sampling-rate sEMG gesture recognition, the recognition accuracy is greatly improved; the influe