CS Student Presented in Top Security Conference

Linghan Zhang, a second year PhD student, has been conducting research in mobile security as her dissertation topic under the guidance of Prof. Jie Yang. Her recent work on voice liveness detection has been published in a paper, titled “VoiceLive: A Phoneme Localization based Liveness Detection for Voice Authentication on Smartphones” in ACM CCS 2016, a top tier conference in the Security area. The paper is co-authored with her lab-mate Sheng Tan, Prof. Jie Yang, and Prof. Yingying Chen at Stevens Institute of Technology.

In the paper, Linghan proposed VoiceLive, a practical liveness detection system for voice authentication on smartphones. It works by leveraging the user’s unique vocal system and the stereo recording of smartphones. It performs liveness detection by measuring TDoA changes of a sequence of phoneme sounds from the two microphones of a smartphone. Experimental evaluation shows that VoiceLive achieves over 99% detection accuracy at around 1% Equal Error Rate (EER). Results also show that VoiceLive is robust to different phone placements and is compatible to different sampling rates and phone models.