Sajda Creates Brain-Computer Interface
"The human brain is perhaps the most fascinating and complex signal processing machine in existence," says Paul Sajda, associate professor of biomedical engineering. "It is the ultimate visual processor." Using expertise in signal processing, computer vision and neuroscience, Professor Sajda has created a "cortically-coupled computer vision system," known as C3 Vision, a brain computer interface system that makes searching through large volumes of imagery more efficient.
When the brain has a flash of recognition--an image it has seen before or an object it is looking for--it emits a signal that can be detected by an electroencephalogram (EEG). When EEG electrodes are attached to a subject's scalp (see illustration above), the system can monitor changes in the brain's electrical activity as the subject watches a video that runs at 10 times its normal speed.
"We are able to detect the neural correlate to the conscious recognition of an object," says Professor Sajda, "the equivalent of the ‘aha' moment." A change in the neural signature indicates a flash of recognition. When detected by the new technology, the images are flagged and reviewed more carefully.
This technology may be used to quickly review hours of surveillance tapes for face recognition or signs of suspicious activity. It also might have applications for review of visual medical information, such as X-rays or MRIs.
Support for developing and testing C3Vision technology comes from DARPA. Professor Sajda is Principal Investigator for a $2.6 million grant, with Truman Brown, Hudson Professor of Biomedical Engineering and professor of radiology, and Prof. Shih-Fu Chang of Electrical Engineering as co-PIs.
Professor Sajda's work will be featured on an upcoming History Channel documentary this fall. You can see how C3Vision works by viewing the video on IEEE's Spectrum website.
Posted:Sep 2 2008