We propose an efficient hand gesture recognition (HGR) algorithm, which can cope with\ntime-dependent data from an inertial measurement unit (IMU) sensor and support real-time learning\nfor various human-machine interface (HMI) applications. Although the data extracted from IMU\nsensors are time-dependent, most existing HGR algorithms do not consider this characteristic,\nwhich results in the degradation of recognition performance. Because the dynamic time warping\n(DTW) technique considers the time-dependent characteristic of IMU sensor data, the recognition\nperformance of DTW-based algorithms is better than that of others. However, the DTW technique\nrequires a very complex learning algorithm, which makes it difficult to support real-time learning.\nTo solve this issue, the proposed HGR algorithm is based on a restricted column energy (RCE)\nneural network, which has a very simple learning scheme in which neurons are activated\nwhen necessary. By replacing the metric calculation of the RCE neural network with DTW distance,\nthe proposed algorithm exhibits superior recognition performance for time-dependent sensor data\nwhile supporting real-time learning. Our verification results on a field-programmable gate array\n(FPGA)-based test platform show that the proposed HGR algorithm can achieve a recognition\naccuracy of 98.6% and supports real-time learning and recognition at an operating frequency of\n150 MHz.
Loading....