Release Date: 2006-11-13
Original Link: http://presszoom.com/story_121144.html
Going through airports or getting sick in a foreign country without speaking the local language can be, at best, inconvenient and, at worst, dangerous. A portable, virtually instant, two-way translator would obviously make all the difference.
(PressZoom.com) - Yuqing Gao, manager of speech recognition and understanding for IBM Research, saw that clearly. So she and her team took up the challenge. And a challenge it was.
“Look at recognition alone,” she said. “There are different dialects, accents, some low voices, some high pitched, older voices, a speaker who uses slang or who slurs or who speaks slowly. And, no matter what, there is always background noise.”
Gao and her team worked to add semantic meaning and context to existing speech recognition. To work as intended, the system needed to recognize English and translate it to Arabic or Chinese and then recognize Arabic or Chinese and translate it back again. But Arabic dialects, for instance, are mainly oral, not written. And the written language is very different from the spoken language. All those issues had to be dealt with. In response, the team invented algorithms and filed more than 10 patents specifically related to these problems.
They also aimed for a tool that could work with limited physical resources.
“We had to shrink and optimize the algorithms to make sure the application could run on a very small device. We optimized the code for laptops, tablet PCs and handheld devices. The target is to optimize it further to work on smart phones,” said Gao.
The result is MASTOR (Multilingual Automatic Speech-to-Speech Translator).
Gao sees the future of MASTOR in smaller and smaller devices, and in more languages. MASTOR can manage Mandarin Chinese and Arabic -- complex languages -- pretty well.
“Now that those are done,” Gao said, “the European languages will seem easier to tackle.”
MASTOR is already in a trial run with the U.S. military in Iraq, but it holds huge promise for business, providing instant translations for videoconferences, conference calls and call centers.
Learn more:
MASTORing languages
http://domino.research.ibm.com/comm/wwwr_thinkresearch.nsf/pages/20061110_mastor.html