SRPeek: Super Resolution Enabled Screen Peeking Via COTS Smartphone.

Jialuo Du,Chenning Li,Zhenge Guo,Zhichao Cao

DOI: https://doi.org/10.1109/icpads53394.2021.00117

2021-01-01

Abstract:The screens of our smartphones and laptops display our private information persistently. The term “shoulder surfing” refers to the behavior of unauthorized people peeking at our screens, easily causing severe privacy leakages. Many countermeasures have been used to prevent naked eye-based peeking by reducing the possible peeking distance. However, the risk from modern smartphones with powerful cameras is underestimated. In this paper, we propose SRPeek, a long-distance shoulder surfing attack method using smartphones. Our key observation is that although a single image captured by smartphone cameras is blurred, the attacker can leverage super-resolution (SR) techniques to recover the information from multiple blurry images. We design an end-to-end system deployed on commercial smartphones, including an innovative deep neural network (DNN) architecture, StARe, for efficient multi-image SR. We implement SRPeek in Android and conduct extensive experiments to evaluate its performance. The results demonstrate we can recognize 90% of characters at a distance of 6m with telephoto lenses and 1.8m with common lenses, calling for the vigilance of the Quietly growing shoulder surfing threat.