Development of a speech interface for control of a biped robot

Abstract

In future it is expected that speech recognition systems to be used as the main man-machine interface for robots used in rehabilitation, entertainment etc. This paper describes the development of a hidden Markov model based speech recognition system for the control of a biped robot. The complete system consists of three subsystems, the speech recognition system, a central controller and the robot. The robot used as the experimental platform was a statically stable biped robot due to its ease of control. The speech recognizer was trained using a set of English words related to biped locomotion. The words chosen were combined to produce meaningful sentences, which after being recognized, were converted to motor commands by the central controller. Two experiments were conducted in which subjects used speech commands to move the robot in a straight line and then in a curved path. The results prove that it is possible to control a biped robot for locomotion using speech commands.