AT&T Developing iPhone Speech Apps

AT&T, the exclusive carrier of the Apple iPhone in the United States, is  developing its own voice applications for the mobile device. Research is also ongoing surrounding a similar program for the BlackBerry and other IP-enabled mobile devices.

The voice controls are being developed in AT&T's labs as part of the phone carrier's Watson Speech Mashups, a speech software framework that casts voice recognition technology as a Web service that can be accessed from anywhere via a high-speed Internet connection. The server collects the voice commands, interprets them into programming language, and relays them back to the device or application to be acted upon, according to AT&T officials.

AT&T also says that by offering speech recognition as a mashup it can open its technology to a broader range of devices and applications since they run as network-hosted speech technologies. Giuseppe DiFabrizio, lead member of the research staff, states that the mashups will work for multimedia devices with broadband access, including IPTV set-top boxes and smartphones like the BlackBerry and iPhone, without the need to install, configure, and manage speech recognition software and equipment.

With the mashup, speech "becomes a viable solution on any IP-enabled device with a broadband connection," Fabrizio explains.

Though AT&T has not set a time line for when and if the mashup will be commercially available, an AT&T research video released on AT&T Labs’ Web site shows an individual using an iPhone to access information from the Yellow Pages mobile Web site, entering city and business requests into the search engine by voice.

Though originally rumored not to support applications native to the iPhone, Fabrizio debunks those statements. "The general idea is to try to move speech to a Web service," he explains. "The user opens a channel that grabs the speech application from the device and runs it through our server."

And, because the application is Web-based, other users and third-party developers can also make use of the service, he adds. Another advantage is that housing the applications on servers can allow for natural language and larger vocabularies and grammars, something often not possible on embedded applications given the mobnile devices' often limited amounts of memory.

In addition to the AT&T applications, Fonix Speech announced that it developed a voice-activated application for the iPhone 3G that could be downloaded directly to the iPhone. Called iSpeak, the Fonix client-side application allows iPhone users to place calls, access contact lists, navigate and access their music libraries, and perform several other functions, all with their voices.

Fonix announced the launch of iSpeak July 7, claiming it as the first run-time application allowing users to interact with the personal contents of their iPhones.  Unlike other voice applets that enable voice search of the Internet by sending commands over the airwaves, this client-side application gives users the power of voice interaction with their personal content and eliminates network latency. 

A few weeks earlier, vlingo released a similar application for the BlackBerry that lets users compose and send email and text messages, search the Web, open applications, place calls, and look up contacts, all with their voices.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues