iType: Using eye gaze to enhance typing privacy
Authors:
- 1 May 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-9
Abstract
This paper presents iType, a system that uses eye gaze for typing private information on commodity mobile platforms. The design combats three primary challenges: 1) relatively low accuracy of mobile gaze tracking; 2) difficulties in correcting input errors due to lacking the comparison with the true text-entry value; and 3) device motions and other noises that may interfere gaze tracking accuracy and thus the iType performance. We devise a set of effective techniques, including leveraging a collective behavior of the gaze tracking results, unique correlation of the typing error spatial distributions, and motion sensor hints from mobile devices, to address above challenges. A set of enhancement techniques are applied to further improve iType's robustness and reliability. We consolidate above designs and implement iType on iOS platform. Evaluations show that iType achieves high keystroke detection accuracy for the secure typing within a reasonable short latency.Keywords
This publication has 22 references indexed in Scilit:
- Snooping Keystrokes with mm-level Audio Ranging on a Single Phone2015
- CIDER2015
- Keystroke Recognition Using WiFi Signals2015
- Tracking Keystrokes Using Wireless Signals2015
- Privacy-Aware High-Quality Map Generation with Participatory SensingIEEE Transactions on Mobile Computing, 2015
- Ubiquitous keyboard for small mobile devices2014
- Tapprints2012
- Biometric-rich gestures2012
- Student's t Distribution2010
- Password hardening based on keystroke dynamicsInternational Journal of Information Security, 2002