Abstract :
Voice communication with computers provides many advantages, particularly for communicating with multifunctional spaceborne computers. Advantages include proper utilization of the astronaut´s (or any computer user´s) training and experience with speech, advantages of natural-language man-computer communication, physical mobility permitted to the busy astronaut, reasonable communication capacities of the voice channel, advantages of multimodal communication, including increased reliability of I/O systems, possibility of monitoring astronaut emotional state, the small closed spaceborne environment with no outside noise and limited population of speakers, and the potential for voice I/O under various accelerations. In contrast, difficulties of oral communication with computers include interference of auditory noise, lack of any direct permanent record of what has been spoken, difficulties of computer handling of versatile natural languages, and increased demand for computer storage and processing capabilities. A system for experimentally comparing speech parameter extraction techniques has been implemented and should provide valuable results concerning the potential for successful speech recognition by computer. The final test of the value of the oral man-computer communication link will be whether human users accept and use it with some ease. Behavioral investigation of the operational acceptability of voice I/O under various restrictions on the communication and command languages used will provide precise measures of the potential success of restricted voice I/O systems.
