37 - Voice processing

Abstract

Voice processing is the term used for the technologies of storing and replaying speech, compressing and decompressing speech, and for the technologies of speech recognition and text-to-speech conversion. Voice processing products can be split into the following types: audiotex, voice response, voice mail, and automated attendant. Audiotex systems are used to publish spoken information to callers. Callers can interact with audiotex systems using dual-tone multi-frequency (DTMF) tones, dial pulses, by speech power detection or in more advanced systems by using speech recognition. Voice response systems are used to allow callers access to databases using a telephone. These systems require the callers to specify the data they are looking for and so intrinsically are interactive services requiring the use of DTMF or voice recognition techniques. Fax response systems, when incorporated within a voice processing system, have two basic modes of operation; store and forward on demand and the on-line transmission of information constructed for the particular caller from a database. The use of automated call handling is becoming important for large corporations striving to reduce costs, make their services available 24 hours a day, and improve their image.