Brief Announcement: Deriving Context for Touch Events
To quantify the amount of high-level context information which can be derived by observing only a user’s touchscreen interactions, we performed a user study, in which we recorded 160 touch interaction sessions from users running different applications, and then applied both classical machine learning methods and deep learning methods to the results. Our results show that it is possible to derive higher-level user context information based on touch events alone, validating the efficacy of touch injection attacks.
KeywordsMachine learning Malicious hardware Smart phone
- 1.Motorola Mobility. Cracked screens and broken hearts - the 2015 motorola global shattered screen survey. https://community.motorola.com/blog/cracked-screens-and-broken-hearts
- 2.Shwartz, O., Shitrit, G., Shabtai, A., Oren, Y.: From smashed screens to smashed stacks: attacking mobile phones using malicious aftermarket parts. In: 2017 IEEE European Symposium on Security and Privacy Workshops, EuroS&P Workshops 2017, Paris, France, 26–28 April 2017, pp. 94–98. IEEE (2017)Google Scholar
- 3.Shwartz, O., Cohen, A., Shabtai, A., Oren, Y.: Shattered trust: when replacement smartphone components attack. In: Enck, W., Mulliner, C. (eds.) 11th USENIX Workshop on Offensive Technologies, WOOT 2017, Vancouver, BC, Canada, 14–15 August 2017. USENIX Association (2017)Google Scholar
- 5.Meng, Y., Wong, D.S., Schlegel, R., Kwok, L.: Touch gestures based biometric authentication scheme for touchscreen mobile phones. In: Kutyłowski, M., Yung, M. (eds.) Inscrypt 2012. LNCS, vol. 7763, pp. 331–350. Springer, Heidelberg (2013). https://doi.org/10.1007/978-3-642-38519-3_21CrossRefGoogle Scholar