Thursday, September 16, 2004

IBM Ponies Up Speech Tech

IBM Ponies Up Speech Tech
IBM has handed over speech technology to two open-source groups to boost development of speech-recognition applications, which proponents say will eventually become common in customer-service call centers, automobiles and elsewhere. The Armonk, N.Y., company has donated code called "reusable dialog components" to The Apache Software Foundation, developer of the popular Apache web server. The RDCs help in enabling telephony software, for example, to understand basic words such as dates, times and locations, i.e. cities and states.

In addition, IBM handed over software tools for speech editing to the Eclipse Foundation.

Both donations are valued at about $10 million, said Brian Garr, program director for the call center and voice portal segment of IBM's pervasive systems unit. The process for standardizing the technology within the open-source groups is expected to take seven to eight months.

The motivation behind IBM's giveaways is to speed up the use of speech-recognition software, according to the company.

"People will be able to develop speech applications for less upfront cost and deploy the software faster," Garr said. "We hope to really elevate all speech applications, so there is more deployment in the middleware infrastructure of corporations."

Garr declined comment on whether the move was to outmaneuver a key competitor, Microsoft. The giant software maker has built its own tools for making speech recognition software for the company's Windows and .Net platforms.

In March, Microsoft introduced Speech Server 2004 for running speech-enabled applications. More than 100,000 developers have downloaded the company's free software developer's kit for building speech applications for Microsoft's platforms

Read complete article. . .

No comments: