Microsoft patents voice-based caller ID, GUI of screen phone

Redmond (WA) - Microsoft has been granted a patent, which describes the identification of telephone callers not by a source telephone number, but by voice characteristics instead.

The patent headlined "Automatic identification of telephone callers based on voice characteristics" describes an idea in which Microsoft envisions future phone calls to be routed through "acoustic repository" - a database that stores sound patterns - that could be used as a potential replacement of certain caller ID services.

The repository is used to compare voice characteristics of a current telephone caller with the pattern previous callers - and hopefully find a match. The downside of this system, of course, is that only callers who have called the system before can be identified. New callers, according to the patent, will have to provide a "voice input" in order to be added to the acoustic repository.

Described applications for the technology range from regular conversations between callers to interactions with call centers. While callers often use the option of being "uncooperative" to bypass much of the queues in annoying automatic phone systems, a voice caller ID may be able to route certain callers much faster to destinations they are looking for.

In an unrelated patent grant, Microsoft also secured the rights to the graphical user interface of a screen phone. The layout described includes information about date and time, incoming calls and messages, call logs and voicemail, as well as shortcuts to speed dial, email, notes, phone directory, address book and Internet access.