Visual whitelisting software helps secure video chats

Visual whitelisting software helps secure video chats

Researchers have developed software that helps keep video chats private.

Duke University computer scientists created software that lets users of camera-equipped smartphones, laptops and other devices specify what others can see, preventing inadvertent disclosure of restricted information that’s in a camera’s field of view, according to the university’s DukeToday publication.

The software helps contain the risk of accidental information leaks that happen when confidential information on a whiteboard or nearby computer screen gets picked up by a video camera or papers on a desk show up in the background of a photo an employee takes of a receipt for filing expenses.

“There are more and more cameras every year. They’re incredibly useful,” Landon Cox, an associate professor of computer science at Duke, said in the article. “But the downside is we’re now converting large swaths of our surroundings to a digital format that’s easy to access and share, including things we might not want to be digitizing.”

The Duke researchers built two tools, one to whitelist information on two-dimensional surfaces such as a whiteboard or slide presentation and the other for 3-D ones, such as people or products. With both, users choose the part of a scene that can be shared by drawing a rectangle around it by hand or with a mouse.

“Once it knows what it’s looking for, the software intercepts all incoming frames from the video stream and rapidly scans frame by frame for a match using computer vision technology,” the article states. “Only authorized objects are allowed to pass from the camera to third-party software, like smartphone apps. Everything else is blocked out by default.”

The researchers tested the tools by asking 26 people to use Android smartphones to scan QR codes with and without the security tools and rate the speed and ease of the “secure” smartphones’ scans. They also tested the software on videos taken while the camera was moving and found that areas could remain protected while the camera delivered 24 frames per second, fast enough for human eyes to perceive a smooth video.

In the past, developers used a blacklist approach for live video redaction, meaning they would build software that blurs or masks information they think would not be sharable, the article states, but it’s hard to anticipate everything that might be sensitive. Changing light conditions and motion added to the challenges, too.

“Even if it fails just 1 or 2 percent of the time, it’s not secure,” said Animesh Srivastava, a graduate student at Duke who co-authored the research.

Next on the researchers’ agenda is studying ways to provide similar privacy controls over audio recordings.

About the Author

Stephanie Kanowitz is a freelance writer based in northern Virginia.

Stay Connected

Sign up for our newsletter.

I agree to this site's Privacy Policy.