Phone: +41 44 632 51 64
My interest lies in leveraging large, crowd-generated data for context-aware applications. Towards this, I build systems to explicitly instrument the masses as well as mine implicit cues from web-sized repositories, such as social media platforms. I'm involved in the SmartDAYS project here in the Wearable Computing Lab. In the past, I studied intelligent systems for robotics and heuristic-based optimization techniques.
I completed my undergraduate studies at the University of Waterloo in Systems Design Engineering. Meanwhile, I experienced various internships in the US and Canada at organizations such as Sun Microsystems Laboratories (now Oracle Labs), Defence R&D Canada, and BlackBerry. After graduation, I continued my studies in Computational Science and Engineering at ETH Zurich. During this time, I exchanged at the Hong Kong University of Science and Technology. In September 2011, I completed my master thesis analyzing the YouTube social network at Google Zurich.
SmartDAYS Project: My aim in this project is to fuse together crowd-generated urban semantics with on-body sensing from commercial wearable devices for understanding what activities are conducted where in the urban space. Using in-situ social media reporting and geo-context obtained through mobile GPS signals, we are able to accurately classify 10 classes of activity routines (e.g. Eating/Drinking, Working, Socializing) with an accuracy of up to 80%.
Urban Risk Mapping: Collaborating with the University of Zurich Travel Clinic, we are creating a smartphone-based risk-aware platform to understand and predict risks while people travel. Using self-report data from travellers, we learn the mapping between risk-related behaviours and the urban spaces that afford such risks.
Courses (as TA): Wearable Systems I (HS 2012)
Projects: Below are some available or on-going student projects. Drop me a line if you are interested in working with me on one below or have a proposal of your own.
Visualization of the Puget Sound Region Council Travel Survey 2006. We learn the mapping from various contextualizing cues (who, when, where) to infer people's travel purposes (why). We are able to train an machine learning algorithm to infer with an accuracy of over 75%. For a quick summary, you take a look at my slides for the 2014 Urban-IoT talk.
Diese Website wird in älteren Versionen von Netscape ohne graphische Elemente dargestellt. Die Funktionalität der Website ist aber trotzdem gewährleistet. Wenn Sie diese Website regelmässig benutzen, empfehlen wir Ihnen, auf Ihrem Computer einen aktuellen Browser zu installieren. Weitere Informationen finden Sie auf
The content in this site is accessible to any browser or Internet device, however, some graphics will display correctly only in the newer versions of Netscape. To get the most out of our site we suggest you upgrade to a newer browser.