home · mobile · calendar · colloquia · 2007-2008 · 

Colloquium - Beveridge

Automated Face Recognition at CSU and at Large
Colorado State University

I will briefly survey automated face recognition research in general and provide an overview of current work at Colorado State. Key elements of automated face recognition will be introduced, and I will briefly touch on recent progress in general, as documented in the recently released Face Recognition Vendor Test 2006. I will also review three projects at Colorado State. In the first, we use statistical analysis techniques to measure the influence of factors such as age, gender, facial expression and focus on algorithm performance. In the second we test different features for face detection, and assess whether commonly used features are actually better than simply using pixels. In the third project, we examine the high-dimensional manifold structure of face imagery and relate this to the difficulty of common data sets.

Hosted by Jane Mulligan.

Department of Computer Science
University of Colorado Boulder
Boulder, CO 80309-0430 USA
May 5, 2012 (14:13)