Now Speaking to the Mobile Market: Voice Command

New speech recognition software for PDAs and cell phones is hailed as the the perfect tool (or toy) for commuters and enthusiasts -- but the possibilities for disabled workers are just as promising.
Posted November 3, 2003

Jim Wagner

For PDA or cell phone enthuasiasts frustrated by poking a stylus or fingers on seemingly microscopic-sized keypads, Microsoft's Voice Command application might be a welcome change.

The software giant's Voice Command (version 1) was released Monday morning at and for $39.95, for PDA or cell phone users using the Windows Mobile 2003 software on their PocketPC.

The application's speech-recognition recognizes English, regardless of the accent, and connects users to their contact list, calendar or Windows media center.

It's a departure from many other so-called speech recognition applications on the market, in which the software recognizes only key words or phrases, or after synching the speech-recognition engine to the users voice. Voice Command, officials said, is truly plug-and-play.

If a user says, "Call Shirley Jones," the software will immediately start dialing the phone number in the contact list. If there is more than one phone number, a female voice will chime back asking whether you want the work or home number.

When a user says, "play I Can't Get No Satisfaction," the software will load up Windows media center and play the Rolling Stones tune. Likewise, you can ask what your next appointment is and it will reply with the next scheduled event in the calendar.

The application will also open up other applications like Word, calculator or even Solitaire.

The biggest trick for Voice Command developers at Microsoft, said Peter Wengert, a marketing manager at Microsoft, was getting the desktop version of Voice Command scaled down to the current 4-5MB file size.

"It's been a challenge, but we've got some great people who were able to port the speech recognition app down to the embedded platform," he said. "The thing you sacrifice isn't necessarily the speech recognition when you go to an embedded device, it's really the size of the text-to-speech engine. If you want to get all the vocabulary of a naturally-sounding person, you really need the power of a desktop PC. But she does the job."

Initially, Microsoft was marketing this strictly as a tool for the 86 percent of commuters who, according to the National Highway Traffic Safety Administration, who felt using a PDA or cell phone made driving much more dangerous.

Microsoft has been working with the automotive industry since 1995 to collaborate on software that will create a safer environment for drivers; there are vehicles on the road around the world with built-in speech-recognition, Bluetooth synchronization with other wireless devices and navigation.

But after more than 1,000 beta users tested Voice Command (off the highway), officials found people were using it for everyday use, saying the product made life a lot easier.

"It's a great application for the enthusiast, but what we're finding as we reach the mass consumers is what one person said, 'it's great for the lazy people,' " Wengert said. "I have a contact list with more than 200 contacts. It's a lot easier to just say the name instead of doing a search and typing in the first few letters."

One other segment of users is being overlooked by Microsoft marketers however: the millions of disabled consumers and workers in the U.S.

Brewster Thackeray, National Organization on Disabilities spokesperson, said Voice Command opens the doors to many people who really need the technology.

"What's unfortunate many times is that technology becomes available to people but it's not affordable," he said. "What jumped out at me was the price; it's exciting, because we need more high-tech assisted technology that's also affordable."

While the less-than-robust capabilities of the print-to-speech software of Voice Command makes its usefulness for blind people limited, Jim Denham of the American Foundation for the Blind said any technology that helps people with disabilities is a good thing.

"We're all in favor of anything we can do to engage the other senses, vis a vis, speech and auditory output," he said. "But until we can get full speech output, it's not going to help visually impaired people a whole lot."

Wengert said the feedback from its beta testers has provided a wealth of information to include in version 2, one of which includes a more complete print-to-speech element.

Another limitation of the software, its American English-only software, he expects will be expanded with software updates down the road.

"With other languages, we have some work to do," he said, but said some of the other major languages will come out "very soon."

Developers around the world have been working on building voice-enabled applications for mobile devices like PDAs. Earlier this year, the Speech Application Language Tags forum made strides to tying HTML and XHTML applications with mobile devices.

0 Comments (click to add your comment)
Comment and Contribute


(Maximum characters: 1200). You have characters left.