Purdue researchers working to make smartphones smarter

WEST LAFAYETTE, Ind. (WLFI ) – Most people rely on their smartphones to help them get from place to place, organize their schedules and keep them connected with the rest of the world. Assistant Professor of Biomedical Engineering Eugenio Culurciello said researchers at the Purdue Weldon School of Biomedical Engineering are working to take that technology a little farther.

“Our main goal is to develop a technology that would allow cellular phones and mobile computers to understand the world,” said Culurciello.

The researchers call it “deep learning”, the calculation mimics how the human brain processes information. Although Purdue didn’t develop the algorithm, Culurciello said they created the hardware needed to make the deep learning run faster. Which gives them the ability to make a mobile device analyze pictures — tagging them with words.

“With our instruments you’re able to tag every image to give them the name of all the content. Turn them into text, and they can be searched later. It would give them the ability to really understand your data, and preferences, and be able to provide applications that would really help you on a daily basis,” said Culurciello.

Culurciello said the technology could eventually be used to control more than smartphones.

“It would give machines, cellular phones and mobile devices the ability to visually process a scene. For example, a car that drives by itself, because it could automatically detect the road or other cars or other objects,” said Culurciello.

Culurciello said although it may seem like humans are giving more responsibility to computers, he said it allows people to focus on what’s important.

“It’s more freeing us from tedious tasks. Giving us the ability to do other important things that we like,” said Culurciello

The technology is still in development, but within a year researchers plan to have a microchip small enough for use in smartphones.

blog comments powered by Disqus