Speech-based cursor control

Abstract

Speech recognition can be a powerful tool for use in human-computer interaction. Many researchers are investigating the use of speech recognition systems for dictation-based activities, resulting in dramatic improvements in recent years. However, this same experimentation has confirmed that recognition errors and the delays inherent with speech recognition result in unacceptably long task completion times and error rates for cursor control tasks. This study explores the potential of a speech-controlled grid-based cursor control mechanism. An experiment evaluated two alternative grid-based solutions, both using 3-3 grids. One provided a single cursor in the middle of the grid. The second allows users to select a target using any of nine cursors. The results confirm that the nine-cursor solution allowed users to select targets of varying size, distance and direction significantly faster than the one-cursor solution. Overall results are encouraging when compared to earlier evaluations of other speech-based cursor control solutions.