IBM shows off speech applications

IBM has showcased a host of prototype and commercial voice recognition applications built by its research teams and partners.


IBM has showcased a host of prototype and commercial voice recognition applications built by its research teams and partners.

The computing giant has been doing research on speech technologies for about 10 years, but speech-enabled applications have come a long way since the company first introduced dictation technologies to the medical industry, David Nahamoo, chief technology officer of speech technology at IBM Research, said.

IBM's partners have had a big part to play in developing the applications based on the company’s ViaVoice technology, he said. "We're working with partners to make a lot of things happen so we accelerate the delivery of useful products and offerings in the marketplace.”

Opus Research analyst Dan Miller said the company had also accelerated its ability to offer speech-enabled applications that were "truly useful" by integrating speech research into its core software group.

This meant IBM was “willing to bring solutions that involve the full line of WebSphere middleware, and application servers along into solutions that indeed involve speech", he said.

IBM's longstanding tradition of innovation also brought credibility to speech-enabled applications and could inspire more people to use them, Miller added.

Companies including Navigation Pioneer Electronics and All Media Guide have used Embedded ViaVoice speech-recognition technology in commercial applications.

Pioneer's new AVIC-Z2 car navigation system provides detailed driving directions that go beyond the usual "turn left at the next street" instructions with the help of Embedded ViaVoice, the company’s product planning manager, Ted Cardenas, said. "It will actually tell you the street name.”

All Media Guide, which provides information about music, artists, songs and albums for websites, online music stores and other entertainment outlets, has teamed with IBM and Avoca Semiconductor to allow users to interact by voice with their personal media collections, said Zac Johnson, product manager for All Media Guide.

Avoca's speech-enabled user interface, which uses IBM Embedded ViaVoice, is called the Media Control Platform. All Media Guide provides the engine within the platform that lets people use voice capabilities to find artists and songs they want to listen to, he said.

For example, a user can tell the system it wants to hear Bruce Springsteen, or tell it to "play The Boss", using the singer’s nickname, Johnson said. The technology can work with a variety of devices on which users store their digital media, such as PCs, iPods or MP3 players.

But Miller said IBM’s big competitor, Microsoft, was also moving into the voice technology space. Microsoft is focused on embedding voice into its Office Communication Server collaboration software, he said. This would provide interesting competitive offerings down the road as IBM is doing the same with its Lotus/Sametime collaboration suite, Miller said.

"I think those companies are really pushing each other to innovate in that area," he said.

"Recommended For You"

Twitter unveils new #music app and service IBM puts Watson supercomputer to work on healthcare