TY - GEN
T1 - Jekyll and Hyde
T2 - 2018 International Conference of the Biometrics Special Interest Group, BIOSIG 2018
AU - Matovu, Richard
AU - Serwadda, Abdul
AU - Irakiza, David
AU - Griswold-Steiner, Isaac
N1 - Funding Information:
This research was supported by National Science Foundation Award Number: 1527795.
Funding Information:
VI. ACKNOWLEDGMENT This research was supported by National Science Foundation Award Number: 1527795.
Publisher Copyright:
© 2018 Gesellschaft fuer Informatik.
PY - 2018/10/10
Y1 - 2018/10/10
N2 - To combat privacy attacks that exploit the motion and orientation sensors embedded in mobile devices, a number of recent works have proposed noise injection schemes that degrade the quality of sensor data. Much as these schemes have been shown to thwart the attacks, the impact of noise injection on continuous authentication schemes proposed for mobile and wearable devices has never been studied. In this paper, we empirically tackle this question based on two widely studied continuous authentication applications (i.e., gait and handwriting authentication). Through a series of machine learning and statistical techniques, we show that the thresholds of noise needed to overcome the attacks would significantly degrade the performance of the continuous authentication applications. The paper argues against noise injection as a defense against attacks that exploit motion and orientation sensor data on mobile and wearable devices.
AB - To combat privacy attacks that exploit the motion and orientation sensors embedded in mobile devices, a number of recent works have proposed noise injection schemes that degrade the quality of sensor data. Much as these schemes have been shown to thwart the attacks, the impact of noise injection on continuous authentication schemes proposed for mobile and wearable devices has never been studied. In this paper, we empirically tackle this question based on two widely studied continuous authentication applications (i.e., gait and handwriting authentication). Through a series of machine learning and statistical techniques, we show that the thresholds of noise needed to overcome the attacks would significantly degrade the performance of the continuous authentication applications. The paper argues against noise injection as a defense against attacks that exploit motion and orientation sensor data on mobile and wearable devices.
KW - continuous authentication
KW - gait authentication
KW - handwriting authentication
KW - wearables and mobile phones
UR - http://www.scopus.com/inward/record.url?scp=85060019310&partnerID=8YFLogxK
U2 - 10.23919/BIOSIG.2018.8553043
DO - 10.23919/BIOSIG.2018.8553043
M3 - Conference contribution
AN - SCOPUS:85060019310
T3 - 2018 International Conference of the Biometrics Special Interest Group, BIOSIG 2018
BT - 2018 International Conference of the Biometrics Special Interest Group, BIOSIG 2018
A2 - Bromme, Arslan
A2 - Uhl, Andreas
A2 - Busch, Christoph
A2 - Rathgeb, Christian
A2 - Dantcheva, Antitza
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 26 September 2018 through 28 September 2018
ER -