The effects of concurrent motor tasking on performance of a voice recognition system

Abstract

This research investigated the effects of concurrent operator motor loading on performance of a voice recognition system comprised of a human operator and a discrete utterance voice recognition system. Increased concurrent operator motor workload (with respect to that experienced during training of the voice recognition system) was found to degrade system performance. Operator motor loading was manipulated using a rotary pursuit tracker. A special vocabulary was used to ensure a baseline recognition error rate to facilitate detection of factors influencing system performance. The results using the special vocabulary also indicated the performance degradations that a real world operator may encounter when using different phrases that are similar to one another in sound.