SRC-Led Research Helps Decode Possible Threats to National Defense, Provides Homeland Security Tools
New Speech-Recognition Technology Recognizes Real-Time Voice Streams Up to Ten Times Faster
RESEARCH TRIANGLE PARK, N.C. - Semiconductor Research Corporation (SRC), the world's leading university-research consortium for semiconductors and related technologies, today announced that researchers at Carnegie Mellon University (CMU) have developed a chip architecture that enables automatic recognition of voice streams at a rate that is up to ten times faster than computers currently supporting speech recognition. The implications of this advancement are exciting the commercial chip industry and are being paid special attention by the defense community.
"The significance of the technology's impact continues to rise," said Dr. David Yeh, director of Integrated Circuits and Systems Sciences at SRC-GRC and a Texas Instruments assignee to the consortium. "Speech recognition has a history of struggling to meet expectations but this research enables dual-purpose applications that were previously untouchable in the consumer space and for purposes of combating security threats to the country."
The research has been targeted at creation of a radically new silicon chip architecture that is exclusively dedicated to speech recognition and performs it vastly more efficiently than current state-of-art computers.
"Whether running an enterprise-class voice call-in center or decoding individual words on a cell phone, all of today's serious speech recognizers exist as software running on some processor. That's terribly limiting," said Professor Rob A. Rutenbar, who leads the In Silico Vox project at CMU. "Moving these computations directly into silicon means we can perform recognition dramatically faster, cheaper, and better for both commercial and homeland security tasks."
As more voice and video digital media goes online, technology users face the increasing challenge of how to search these "live streams" effectively. This emerging "audio mining" problem is especially acute in homeland security applications that need to expedite translations of huge streams of incoming data, sifting for a few critical words or phrases. Fast-recognition technology being perfected at CMU is aimed at making it possible to rapidly search massive volumes of data.
In its third year of a research initiative among the SRC, Defense Advanced Research Projects Agency (DARPA), U.S. National Science Foundation and CMU, the multi-purpose technology will be used by designers of semiconductors, systems and software. Radical cost savings will result from the new chip designs that are also about ten times smaller than the microprocessors in computers currently running speech recognition. This efficiency will help to significantly advance both national security and commercial applications.
Today's announcement reflects rigorous research sponsored by SRC-GRC's Integrated System Design Thrust, SRC-FCRP's C2S2 Development Center, DARPA and the National Science Foundation.