Voice recognition: Same old tune?

Summary:For handhelds, the 90s has been the decade of the stylus. But the next-generation of Pilots, cell phones, and other info gadgets won't need to be constantly prodded, say some.

For handhelds, the 90s has been the decade of the stylus.

But the next-generation of Pilots, cell phones, and other info gadgets won't need to be constantly prodded, say some. Rather, with new voice-recognition technology coming out, the new devices will listen.

"The call for voice recognition is much louder," said Jan Abelev, director of marketing for voice-recognition chip maker DSP Group Inc. (DSPG) "During this year's Consumer Electronics Show, there was much more interest (than in previous years)."

The Santa Clara, Calif., company has bet its future on giving info gadgets the ability to listen. On Wednesday, it announced a voice recognition chip for creating miniature cell phones that dial without the use of a keypad, car navigation systems that listen and answering machines that can interact with the caller.

The technology was homegrown in the company's Israeli development center. "We have definitely improved this technology over the years," said Abelev.

And the market seems to be waiting. Interest in voice recognition has reached critical mass, setting the stage for new products. "It is a very important time to come out with this technology," said Shannon Pleasant, communications market analyst for research group In-Stat. "Voice recognition has matured to the point that devices that use it are actually useful."

While a small part of the programmable digital signal processing, or DSP, chip market -- predicted to reach $9 billion by 2001 by In-Stat analyst Joyce Putscher -- voice-recognition DSPs' share will grow.

Still, many analysts have not yet become enamored with the nascent market. "The hardware itself has been around for some time," said Dean McCarron, principal analyst of market researcher Mercury Research Inc. "The algorithm that do the processing have been fairly well known."

In addition, many technical problems haven't gone away. The new chip's recognition accuracy is about 95 percent -- good, but not great. Other problems include accurately recognizing the speech of different speakers. "Speaker independent recognition, which is notoriously difficult, will limit the application of the technology," said McCarron.

Still, admits McCarron, while the chip may be a small step forward for the voice recognition market, it is still important. "When you put the functions into a chip -- rather than software," he said, "you make the features available to an entirely new market."

The DSP Group's chip integrates telephone features for dialing and device control with the an ability to recognize words and should be available in early spring.

Topics: Hardware

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Related Stories

The best of ZDNet, delivered

You have been successfully signed up. To sign up for more newsletters or to manage your account, visit the Newsletter Subscription Center.
Subscription failed.