Abstract
In this paper, we discuss how to recognize gesture patterns of a human. It is very natural and useful for a human to use a gesture in order to give the robot a specific task. Furthermore, the gesture plays an important role in interaction and communication with human. Therefore, we propose gesture recognition methods of a partner robot for natural communication with human. First, we apply a steady-state genetic algorithm to detect a human hand. Next, we apply a spiking neural network for extracting a human hand motion and a self-organizing map for extracting human hand motion pattern. Finally, we show experimental results of gesure recognition for a partner robot, and discuss the effectiveness of the proposed methods.
