E-mail to voice interpreters getting even more fancy

From...

May 26, 2000
Web posted at: 10:37 a.m. EDT (1437 GMT)

by Mubarak Dahir

(IDG) -- Despite knowing she had important e-mail to read, Nancy King dashed out of her office at 4:50 p.m. on April 12. King, director of research at Lee Koehn Associates, an executive recruiting firm in Portland, Ore., had to pick up her two sons from school, run home, whiz through McDonald's, then dash to the Parkrose Little League baseball field.

	MESSAGE BOARD
Hardware innovation

Once there, King sat in the bleachers, pulled out her cell phone and checked her e-mail.

She dialed Etrieve, a service that accesses customers' e-mail by phone. After she punched in her password, a computer-synthesized female voice said she had one priority message and four regular e-mail. After about 10 minutes of listening to and answering her e-mail, King turned off her phone and watched her son play ball.

Mike Maerz and Ash Gupte, former Intel executives, founded Etrieve in 1998 so business travelers could access e-mail on the road. Etrieve, which made its service available to the public last month, is one of several companies developing e-business applications for text-to-speech, or TTS, technology programs that can "read" then "speak" text through a synthesized human voice.

MORE COMPUTING INTELLIGENCE

IDG.net home page

The Standard.com

TechInformer: The Thinking Internaut's Guide to the Tech Industry

Voice-activated Web browsing standard

New voice portals trumpet the arrival of Internet-via-phones

Reviews & in-depth info at IDG.net

E-BusinessWorld

Industry Standard email newsletters

Questions about computers? Let IDG.net's editors help you

Industry Standard daily Media Grok

Search IDG.net in 12 languages

News Radio

Fusion audio primers

Computerworld Minute

Some companies apply TTS technology to their Web pages to set up office conferencing systems and audible chat rooms or to enhance customer service. But the most common application is e-mail over the phone.

"We're just seeing the industry take its first baby steps" with products that use TTS technology, says Dana Thorat, a senior analyst with research group IDG in Framingham, Mass. Still, she agrees with those who believe the industry may be on the cusp of a TTS products and services explosion. "It will be a big summer," she says.

TTS-based products ö mostly reading devices for the blind ö have been around since the early 1980s. In the early 1990s, according to Janet Pierrehumbert, a professor of linguistics at Northwestern University, a research breakthrough in TTS technology transformed the way sound could be captured and stored.

But its progress stalled: TTS systems were intelligible, but they did not sound natural. (Etrieve's e-mail reader sounds like Rosie, the robot housekeeper on The Jetsons.)

Jim Glass, principal researcher with the Spoken Language Systems Group at MIT's Lab for Computer Science, says TTS research stagnated because "it was always the ugly stepsister of voice recognition." Many researchers considered voice recognition a more challenging field, and the market demand for TTS-driven products seemed limited.

Two technical advances in TTS synthesizers have driven new applications, says Ron Benson, founder and CEO of Vocal Point, a San Francisco company developing a service for Internet access via telephone. One advance is in "subwords," or the tiny bits of recorded speech that form the basis of TTS synthesizers. The other advance is the techniques for sewing together those bits to make words and sentences. Recordable subwords have gotten bigger, and the methods of splicing them together have become smoother, resulting in much more natural-sounding voices.

Controlling pitch, intonation, emotion, stress and tone are only a few of the remaining challenges. And though TTS technology is at the point where it can be understood, Pierrehumbert warns that "after a while, unless the technology keeps pace, people may find it exasperating to listen to."

Even if TTS technology doesn't improve by leaps and bounds in the near future, most observers agree it has a growing place in the market. IDG's Thorat believes the current level of TTS-dependent products can win acceptance if they focus on the right niche markets. "No one's going to want to sit through a reading of War and Peace with these things," she says.

WAP all your messages with Onebox.com
March 22, 2000
Talking around the Web becomes a reality
September 14, 1999
Top 10 multimedia tools for your e-mail
November 17, 1999
You've got VizMail!
July 28, 1999
Net2Phone upgrades PC-to-phone service
December 7, 1999

TalkingNets introduces itself as a voice ASP
(Network World Fusion)
WAP service juggles all messages
(PC World)
Let your messages follow you around
(PC World)
Voice portals: To hear more ads, press one
(The Industry Standard)
Polycom to unveil VoIP conference phone
(IDG.net)
New voice portals trumpet the arrival of Internet-via-phones
(The Industry Standard)
IT finds its voice as telcos pitch voice/data convergence
(InfoWorld)
Standard completed for voice-activated Web browsing
(Network World Fusion)

Etrieve

Note: Pages will open in a new browser window

External sites are not endorsed by CNN Interactive.