NEW YORK — IBM (Quote) thinks that speech technology has
evolved from the first rudimentary dictation products it introduced 10
years ago to mainstream and, in some cases, life-altering applications.
To prove the point, the company introduced a sampling of everyday
applications at its Speech Technology Innovation Conference here.
These include in-car products using voice recognition to improve hands-free
performance of navigation and entertainment features and call center
IBM also demonstrated two applications with important geo-political implications: MASTOR, a speech-to-text-to-speech translator which it deployed in Iraq last year, and TALES, which translates and transcribes foreign-language TV broadcasts and Web sites. TALES currently translates
Arabic and Mandarin Chinese into English.
David Nahamoo, chief technology officer for speech technology at IBM
Research, said that speech technology will prove as important to the world
as telephony and transportation.
The latter technologies “bridged time and
distance. Speech technology will help bridge the cultural divide,” he told
One example of how this could play out is MASTOR.
The tool, which runs on a Panasonic ToughBook, allows U.S. soldiers to speak into a microphone in English and have their words translated into Iraqi Arabic. Their words are also shown on the laptop screen in both English and
Arabic so that the speaker can be sure his or her words were properly
MASTOR also translates from Iraqi Arabic into English, allowing
U.S. soldiers and Iraqi civilians to actually carry out a conversation.
Nahamoo says technological innovations like MASTOR can help people break
language barriers that are at the root of most of the world’s problems.
He noted that IBM Research has the luxury of working with a $600 million
He said the company’s philosophy is to fund projects without preconceptions
and see what commercial applications can be developed later.
“Our approach is, let a thousand flowers bloom,” he said.
But the company also has more tangible goals in mind for speech technology.