USABLE TTS FOR INTERNET SPEECH ON DEMAND

“Speech on demand” applications need to communicate knowledge from a large and/or changing collection of data, and thus can’t pre-record all of the possible verbal information needed. Instead these applications must use some form of synthesis to convert data into an auditory form. Textual data can be transformed into speech through the use of Text To Speech synthesis, and there are certain factors such as bit rate, quality, expression, and delay that impact upon the usability of such speech on demand systems. This paper discusses different implementations of speech on demand systems, using Text To Speech synthesis, and assesses how usability is affected.