A Speech Recognition Client-Server Model for Control of Multiple Robots

摘要

Considerable amount of research has been carried out in the domain of man-machine interaction. Interaction with machines using hand gestures, eye motion, etc. has already been proposed by researchers all over the world. However, interacting with devices using speech is of particular interest to researchers since speech is the most natural way of interaction and communication for human beings. In this paper, we have tried to develop a client-server based architecture for controlling several robots simultaneously through voice commands. The robots used in the experiment are the LEGO® Mindstorm® NXT robots. The entire architecture is developed using the client-server model of communication which enables each and every component of the architecture to be present on a physically different machine and at a physically different location allowing the user to control multiple robots using speech on the go. The speech recognition server accepts speech inputs from the client, translates them into robot control commands and sends them to the specific server controlling the robot. Thus the user need not be physically present at the same location as the robot and is able to control robots remotely.