A GESTURE RECOGNITION SYSTEM USING A PHYSICAL SENSOR BASED ON ARTIFICIAL NEURAL NETWORK
In this paper, we propose a gesture recognition system using a flexible epidermal tactile sensor by using artificial neural network (ANN). We define five different hand motions and obtain those five hand gestures by using flexible epidermal tactile sensor. Artificial neural network is applied to classify the gestures. The result shows that our system using raw data with ANN has a better performance than the conventional machine learning algorithms like SVM using integral absolute value (IAV), difference absolute mean value (DAMV) features. Also, it has a better performance compared to using micro-Doppler signature with deep convolutional neural network (DCNN).
gesture recognition, artificial neural network, hand gesture, flexible epidermal tactile sensor.