Abstract
In this paper, a voice recognition function of the housing VR system, developed as an expansion of the man-machine interface, is reported. As voice recognition software, UNMAGIC, in which the voice recognition rate was very high, was adopted. Voice recognition and VR display were realized by 2 personal computers, when voice recognition and VR display were carried out by stand-alone personal computer, because the recognition rate tended to decline. Thus, a technique that communicated only information for operation between 2 personal computers was proposed in this paper. In addition, from evaluation experiments, the optimum operation width differed according to the transfer direction, and again, the misrecognition tended to fluctuate by property of subjects, and contents of speech sounds were obtained.
