BlackHat资料us-14-Fu-My-Google-Glass-Sees-Your-Passwords-WP.pdfVIP

My Google Glass Sees Your Passwords! (Black Hat USA 2014 White paper) 1 2 1 1 3 4 Qinggang Yue , Zhen Ling , Xinwen Fu , Benyuan Liu , Wei Yu and Wei Zhao 1 University of Massachusetts Lowell, USA , xinwenfu,qye,bliu@ 2 Southeast University, China , zhenling@ 3 Towson University, USA , wyu@ 4 University of Macau, China , weizhao@umac.mo Abstract. In this white paper, we introduce a novel computer vision based attack that automatically discloses inputs on a touch enabled device while the adversary cannot see any text or popup in a video of the victim tapping on the touch screen. We carefully analyze the shadow formation around the fingertip and apply optical flow, Deformable Part-based Model (DPM) based object detector, k-means clus- tering, and other computer vision techniques to automatically locate the touched points. Planar homography is then applied to map the estimated touched points to a reference image of software keyboard keys. Our work is substantially differ- ent from existing research efforts on blind recognition of touch inputs. We target passcodes/passwords where no language model can be applied to correctly esti- mate touched keys. We are interested in scenarios such as conferences and similar gathering places where a Google Glass, webcam, or smartphone can be used for stealthy attacks. Extensive experiments are performed to demonstrate the impact of thi