报告题目: Faces are Protected as Privacy: An Automatic Tagging Framework against Unpermitted Photo Sharing in Social Media

报告人：Sheng Wen 博士，Deakin 大学，澳大利亚

时间: 2017年10月28日 11:00-12:00

地点: 文津楼三段 622报告厅

Abstract: In social media such as Facebook, the sharing of photos among users is common and enjoyable but also very dangerous when the uploader posts photo online without the consents from other participants in the same photo. As a solution, recent research has developed fine-grained access control on social media photos. Every participant will be tagged by the uploader and notified through internal messages to initialise their own access control strategies. The appearances of participants will be blurred out if they want to preserve their own privacy in photo. However, these methods highly depend on the uploader's reputation in tagging behaviours. Adversaries can easily manipulate unpermitted tagging processes on and then publish personal/group photos, which should have kept confidential to the public in social media. In order to solve this critical problem, we propose developing a participant-free tagging system for social media photos. This system excludes potential adversaries through automatic tagging processes over three cascading stages: 1) an initialisation stage will be applied to every new user to collect his/her own portrait samples for future internal searching and tagging; 2) when samples are incapable of tagging a participant in photo, public search engines (Google) will be employed for extensive sample collection; 3) the remaining untagged participants will be identified cooperatively through tagged users complying with Byzantine Failure Principle. In the evaluation, we first demonstrate the security of the proposed system through on-site empirical tests with 25 volunteers on the above three stages. The results suggest that our system is immunised to unpermitted tagging and non-tagging (100% unsuccessful rate). We also carried out a series of experiments to validate our system's efficiency and effectiveness. The results demonstrate the tagging efficiency (91.8% tagging rate), photo processing efficiency (0.86s/photo on average), and adoption willingness (92% in 200 volunteers agree to use our system).

.

报告人介绍： Dr. Sheng Wen was a joint PhD candidate of Central South University and Deakin University (12/06/2011 ~ 30/03/2014). His research interests include modeling of the propagation and defense study of Internet malicious information. He is also interested in the techniques of identifying information origins in networks. His advisors are Prof. Wanlei Zhou and Dr. Yang Xiang at Deakin University, and Prof. Weijia Jia at Central South University. From 01/04/2014 ~ 18/01/2015, he worked with Prof. Ivan Stojmenovic as Research Fellow in Deakin University, Australia. From 19/01/2015, he became a lecturer in computer science in Deakin University. From Oct. 2017, he works as a senior lecturer in Swinburne University of Technology.