LSU Computer Science Professor Developing Software to Shield Against Visual Hacking
October 7, 2020
BATON ROUGE, LA – Ever find yourself looking over your shoulder while unlocking your phone, texting or replying to an email? When your family or close friends borrow your phone, are you worried they may browse private folders without your permission? LSU Computer Science Assistant Professor Chen Wang is ready to put a stop to this by creating a software-based solution for mobile devices to prevent visual hacking.
Traditional defense methods against visual hacking on personal devices, such as smartphones and tablets, are privacy screens, which allow just a splinter of light to pass through within a narrow angle in front of the screen. However, by filtering out light, the film unavoidably dims the visible output and reduces the display resolution. Recent smartphone apps allow the user to generate color pattern overlay to obscure the screen contents, but this also significantly affects the user experience.
Most importantly, both of these methods have limited capabilities to prevent visual hacking.
Wang’s proposal is three-fold. First, he plans to develop unobtrusive user authentication methods based on acoustic sensing to confirm the user identity before displaying any information.
“I have submitted several papers already on identifying who is holding the phone because people have different hand geometry, like different size fingers,” Wang said. “Depending on who’s holding the phone, we can release different privileges. It could be in ‘host’ mode.”
This method eliminates the need for facial ID recognition or a password that requires the user to proactively present a face or enter memorized characters.
The second part of Wang’s proposal focuses on enabling the device to detect surrounding visual hackers (human snoopers and camera-based hackers) based on WiFi sensing and deep learning. Wang will create sensing techniques to see if there are other people or cameras around that can zoom in on your screen
“Visual hacking is a high-gain but low-tech attack,” he said. “An adversary could obtain the user’s sensitive onscreen contents by simply stealing a glimpse over the user’s shoulder or covertly taking a snapshot with a camera. The curious peeping eyes or benefit-driven visual hackers exist around us every day, and it is challenging to wipe out such threats.”
The third part of Wang’s proposal entails designing privacy-preserving screen display methods for mobile devices to prevent unpermitted visual captures, leveraging the unique physical properties of liquid crystal (LC) molecules in the screen.
Wang came up with the idea for this project when he was a PhD student. He found students had to turn over their phones on the desk when meeting with their advisors. He figured many others would have the same issue, such as when in a movie theater where the row behind you can see your phone screen.
Wang’s project, which was made possible through a $167,750 Board of Regents grant, is not yet complete. In order to complete his research, he has created a short survey that will collect statistics to help him understand people’s awareness and concerns about smartphone on-screen privacy leakages. To complete the survey and aid in Wang’s research, visit https://www.surveymonkey.com/r/SMPTK9X.
Contact: Libby Haydel