Great strides have been made recently in the areas of speech recognition and speech synthesis. At Apple, recent engineering has resulted in the ability to perform high-quality, real-time continuous speech recognition and synthesis on a personal computer. Casper is a system that integrates speech recognition and synthesis into a spoken interface for the Macintosh (TM) computer. This paper describes this system, and how key aspects of the underlying technology have influenced its design and user interface.

Keywords: speech, speech recognition, speech synthesis, speech interface, human interlace, natural language