Microsoft, Unisys boost speech recognition

Microsoft and Unisys teamed up last week to push speech recognition forward. The alliance will focus on both technical and marketing aspects.

The companies will work to speed the adoption of the speech application programming interface (SAPI) by offering developers a tool set and SAPI-based software.

As a result of the alliance, "a typical Microsoft application developer or interactive voice response developer will be able to create speech recognition applications," said Joe Yaworski, vice president and general manger of the Unisys Natural Language Understanding business initiative.

The companies realize that, in addition to a lack of applications and tool sets, insufficient marketing is another reason that speech recognition has not taken off. "There hasn't been a big marketing group behind [speech recognition]. The fact that Microsoft and Unisys have teamed up behind it should help a lot," said David Jennings, product planner at Microsoft.

Unisys also has launched Speech-Depot, a development Web site, that will supply the components, information and products for developers.

But whether seeding development will let speech recognition software go mainstream is questionable because of other factors, said Rob Enderle, a vice president at Giga Information Group. "It is not just a technical issue but a behavioral issue," he said. This technology may find more acceptance as a Web access tool, Enderle said.

And the companies are looking in that direction. "There is no doubt about it, what we are doing with Unisys will eventually play out on the Web," Jennings said.