InformationWeek
Defining The Business Value Of Technology
November 10, 1997
Mix-And-Match Package For Voice Recognition
API integrates four levels of technology
The Dallas company's SpeechWave 1.0 uses one API to integrate four modules: Fundamental, Standard, Premium, and Secure. Developers can use the most cost-effective and reliable component for different parts of the same application.
"You don't need a 10,000-word recognizer for an application that just asks the speaker to say yes or no," says Jeff Engle, director of product marketing for Voice Control. For that, a developer would use SpeechWave Fundamental, which is designed for applications with just a few words and single-digit numbers. It starts at $300 per port. SpeechWave Standard, at $500 per port, is more reliable for recognizing a long st
ring of numbers in a core business application.
SpeechWave Premium, at $1,100 per port, lets developers build apps with a customized list of about 1,000 words, such as a company's product list. SpeechWave Secure lets applications add speaker verification, which matches the speaker's voice to an electronic copy, also for $1,100 per port.
In the past, users had to turn to several companies, and often to systems integrators, to create single applications that could reliably recognize both numbers and a custom list of 1,000 words. "Anything that simplifies the process is a boon for voice recognition," says Steve Gladstone, general manager of Hammer Technologies Inc., a Wilmington, Mass., company that uses speech-recognition technology in the test equipment it makes for computer telephony networks.
Voice Information Associates, a research firm in Lexington, Mass., estimates that the market for speech-recognition equipment will grow from $180 million in 1996 to $810 million in 2001.
oice-recognition technology could get a lot easier to deploy. Voice Control Systems Inc. last week announced a software developers' package that lets programmers mix and match various levels of speech technology to develop applications that respond to a human voice.
Featured Microsite
