Speech-driven graphics: a user interface

Abstract This paper discusses an ongoing research project concerned with exploring the potential of speech recognition equipment to enable hands-free drawing on the visual display unit of a Macintosh computer system. The paper concentrates on some of the more intractable problems, such as how to control the location and movement of the drawing cursor with any degree of precision, and how to communicate with the recognition software in a way which minimizes perceived ambiguities in the intentions of the user. In particular, consideration of the manner in which graphics commands are verbalized is an essential prerequisite for the successful design of an intuitive use of the system and this is an area on which much of the research effort was concentrated. It is already apparent that it will be very difficult to resolve satisfactorily certain aspects of speech-controlled drawing with the current generation of speech recognizers.