SeSaMe Seminar by Dr Thomas Winkler
Title: Security and Privacy in Visual Sensor Networks
Speaker: Dr Thomas Winkler
Date & Time: 7th August (Wed), 10.30am -12noon
Venue: AMI lab, iCube building Level 2
Abstract: The talk initially motivates the need for security and privacy in video surveillance and visual sensor networks. Smart and embedded cameras open new possibilities to bring security and privacy protection closer to the image sensor. The talk describes the approach of the TrustCAM project to implement such a trustworthy camera system. It integrates low-level system security techniques from the domain of "Trusted Computing" with computer vision based privacy protection executed at the application level. Increasing concerns about widespread video surveillance are addressed by integrating a scheme that enables users with mobile devices to interact directly with cameras to check their status and infer what privacy and security features are implemented. The second part of the talk focuses on recent work in the TrustEYE project which is aiming to design a secure sensing unit that facilitates the separation of sensitive from non-sensitive data within an embedded camera device. The goals, benefits and challenges as well as implementation options are discussed. An early TrustEYE prototype based on an ARM Cortex M4 processor is presented. The system is designed as a research platform which can be used as an add-on for existing embedded development boards or as a standalone camera in a visual sensor network. The talk is concluded with an outlook to future work and topics for collaboration with NUS and the SeSaMe.
Bio: Thomas Winkler studied Telematics at Graz University of Technology and earned his Dipl. Ing. (MSc) with distinction in 2005. His thesis is on mobile agent middleware systems for embedded video surveillance cameras. In fall 2005 he joined the Institute for Applied Information Processing and Communications (IAIK) as a researcher in the EU-funded project "Open Trusted Computing". He worked on Java security solutions and was the initial designer and developer of jTSS - the first Java software stack for "Trusted Computing". He is a member of the JSR321 expert group and the TCG embedded systems work group. In May 2007 he joined the Institute of Networked and Embedded Systems (NES) at Alpen-Adria University Klagenfurt. His research interests focus on pervasive smart camera systems including hardware architectures as well as software aspects such as operating systems and middleware. Security and privacy in visual sensor networks have become central topics of his work. In July 2011 he completed his PhD studies with distinction. As part of his research he developed a TPM-enabled embedded camera system called TrustCAM with inherent security and privacy protection features. In his current work he takes these ideas a step further and develops concepts for a secure sensing unit. One of the goals of the TrustEYE project (grant KWF-3520/23312/35521) is the separation of non-sensitive from sensitive data within an embedded camera.