Ryerson student builds world’s first gesture-based computing database

A PhD student at Toronto’s Ryerson University says he could complete the world’s first extensive database of human gestures for use as a universal computing interface by the end of this summer.

Adrian Bulzacki, now in his final year of Ryerson’s electrical andcomputing engineering PhD program, started working on the universalgestures database while developing a new game called Charades forKinect, Microsoft Corp.’s wildly successful motion-basedgaming system.(Charades is due to be released by Microsoft by the end of thissummer.) He realized while developing that particular game that no one,so far, had put together a comprehensive, catalogued, technicaldatabase of human gestures that can be recognized universally by anygesture-recognition computing system.

“What Google is to online text search,we’re trying to be that forgestures,” said Bulzacki, 29. “We think we’re the first (ones) tocommercialize the data on the gestures.”

Bulzacki’s database collects the best quality gesture recognition datafrom around the world that can be used in computing, then tags thatdata and puts it on a server that is able to learn it and identify it,he explained. His system is more accurate and efficient than others outthere, he said.

“We found our accuracy is unusually high and we thinkit would be highenough to be of military standard,” said Bulzacki, adding that hissystem can recognize gestures 30 times faster than the human eye and,unlike most gesture recognition systems, only requires one humangesture rather than a sequence of several.

Commercial applications
“Now you’ll be able to interact and interface with any environment,” hesaid, not just traditional computer touch screens now used intablets.“It unlocks a door to immersion everywhere, being surrounded by virtualitems and information and interacting with it accordingly,” he said.Bulzacki is also working on a type of wallpaper “to create an immersiveenvironment in your living room.”

The commercial applications could be almost endless, but one exampleBulzacki gives would be to use the system in hospitals so doctors –especially surgeons – can operate computers and machines throughgestures rather than actually touching them, a practice which currentlyforces them to scrub their hands repeatedly to prevent contaminatingpatients and sterile equipment.

Bulzacki has already licensed the system but it won’t be integratedwith a live online recognition network until the end of the summer.

Bulzacki just received $55,000 in funding for his work on the databasefrom the Federal Economic Development Agency for Southern Ontario. Themoney is part of $315,000 in federal funds that will be awarded to 10Ryerson students and graduates inscience, technology, engineering andmath (STEM) to help them commercialize their research and ideas.

Share on LinkedIn Share with Google+