AUTHOR=Zheng Nan , Li Yurong , Zhang Wenxuan , Du Min 

TITLE=User-Independent EMG Gesture Recognition Method Based on Adaptive Learning

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.847180

DOI=10.3389/fnins.2022.847180

ISSN=1662-453X

ABSTRACT=In a gesture recognition system based on surface electromyogram (sEMG) signals, the recognition model established by existing users can not directly generalize to the across-user scenarios due to the individual variability of sEMG signals. In this paper, we propose an adaptive learning method to handle the problem. The muscle synergy is chosen as the feature vector because it can well characterizes the neural origin of movement. The initial train set is composed by representative samples extracted from the synergy matrix of existing user. When the new users use the system, the label is obtained by the adaptive K nearest neighbor algorithm(KNN). The recognition process don’t require the pre-experiment for new users by the adaptive learning strategy, namely, the qualified data and the label of new user data evaluated by a risk evaluator are used to update the train set and KNN weights, so as to adapt to the new users. We have tested the algorithm in DB1 and DB5 of Ninapro databases. The average recognition accuracy is 68.04%, 73.35% and 83.05% for different types of gestures respectively, achieving the effects of user-dependent method. Our study can avoids the re-training steps and the recognition performance will improve with the increased frequency of uses, which will further facilitate the widespread implementation of sEMG control systems using pattern recognition techniques.