EIT Gesture Recognition Method With Stronger Robustness

With the development of computer technology and artificial intelligence, gesture recognition technology has gradually become a research hotspot as an important way of human-computer interaction. Electrical impedance technology (EIT) has the advantages of portability, low power consumption and low cost, which provides a new idea for the design of gesture recognition system. In previous studies, the traditional machine learning algorithm as the classification method of EIT gesture data exposed the limitations of poor robustness and generalization. In this paper, the Convolutional Neural Network (CNN) is used as the feature extraction module to improve the SVM algorithm, and the center loss is used to further enhance the discriminative ability of the classifier. Two interference experiments are designed in this paper. The experimental results show that based on the powerful feature extraction ability of CNN and the supervision of feature compactness by center loss, the new classification method proposed in this paper achieves stronger robustness and generalization ability.