ROCHESTER, N.Y. (WROC) — Some photos and videos you may come across on the web are supposed to look real, even though they’re made in front of a computer screen.
They’re called “deepfakes,” and researchers at the Rochester Institute of Technology are working to develop better ways to detect them.
Researchers are being commissioned by the government as part of a multi-year project to develop an algorithm to catch those fake images.
“We then get the results back to say, ‘this is what was detected, this is what wasn’t detected,'” said Christye Sisson, a professor and researcher at RIT. “So we can actually be better bad guys and exploit some of those weaknesses to help strengthen those algorithms of detection.”
First, they create a fake image or video. Then, they develop tech that allows them to find out whether it is fake or real.
According to Sisson, it always helps to be vigilant on the web.
“The point is really doing that little bit of research and taking the time to find out what the original intent was,” said Sisson.