TABLE OF CONTENTS Enumerating exhaustively all industrial products using speech synthesis would be extremely fastidious. However, it is interesting to sketch the taxonomy of the application domains of this technology.
First, it is essential to recall that a speech synthesis system produces information that is distributed via an acoustic channel. A major feature of this channel is to be rather inadequate for communication multiplexing. In a conversation, the acoustic channel is used almost exclusively by one speaker at a time. The speech synthesis system will turn out to be a good technological choice if the application is guaranteed to be the only one to use the acoustic space available around the information delivery system.
Today, the quality and the intelligibility of synthetic speech is less effective than that of natural speech. Users of information delivery by voice will only accept synthetic speech if they have no other choice.
In summary, the ideal prototype for a service using speech synthesis is a service that must deliver a large quantity of unpredictable information in an ergonomic context where the acoustic channel is as user-friendly as possible.
In the telecommunication sector, several operators propose two leading services based on speech synthesis: reading e-mails on the phone (notably for mobile phones, for which the reading surface is limited) and information delivery concerning telephone numbers (direct or reverse directories).
In the sector of embedded devices, the miniaturization of components makes it possible to design navigation tools for vehicles or more specific systems such as...
