Hidden Camera Detection Using Cyber-Physical Correlation
Prof Brent Lagesse ( University of Washington, Bothell )
- 15:00 3rd July 2018 ( Trinity Term 2018 )Vicon room, CPS group, Level 2, Robert Hooke Building
Talk Abstract
Small, low-cost, wireless cameras are becoming increasingly commonplace making surreptitious observation of people more difficult to detect. Traditional techniques for detecting hidden cameras tend to assume that the user has significant control over their environment; however, this assumption does not always hold. This talk introduces two techniques for detecting streaming Wi-Fi cameras that can be performed in a much wider variety of environments than previously possible. These techniques rely on the ever-growing link between the cyber and physical worlds. We describe techniques for actively manipulating the physical environment and observing correlations in the cyber environment. Next we introduce the concept of similarity of simultaneous observation where the user utilizes a camera (webcam, camera on a phone or laptop) to compare timing patterns of data transmitted by potentially hidden cameras and the timing patterns that are expected from the scene that the known camera is recording. We have demonstrated the suitability of these techniques for detecting streaming Wi-Fi cameras by achieving classification accuracy as high as 97\% accuracy with an F1 score of 0.965 on the diverse data sets we have collected.
Speaker's Bio
Dr. Lagesse received his Ph.D. in computer science from the University of Texas at Arlington in 2009. Prior to coming to UW Bothell, he held positions as a research scientist in the cyber security research groups at Oak Ridge National Laboratory and BBN Technologies. His expertise is in the areas of cyber security and pervasive systems and current projects include building security mechanisms to protect against adversarial machine learning attacks and the development of realistic, dynamically created virtual systems for security and education. Dr. Lagesse's other work involves research into enabling developers to more easily write and deploy secure code in mobile and heterogeneous environments and mechanisms to encourage more secure use of software by users in such environments.
Share this:
