Colloquium "Application of face alignment in driver fatigue monitoring systems"

Abstract:

Face alignment is a procedure in which facial features are aligned to face image. It is used in various computer vision aplications such as face recognition, facial expression recognition, face tracking etc. Over the past several years many methods and algorithms for face alignment are developed, ranging from parameterized models such as active shape and active appearance models to regression-based and deformable models. In this presentation a non-parameterized method of face alignment based on gradient descent was described and applied to driver fatigue monitoring systems.

CV:

Markan Lopar was born 1981 in Sombor, Serbia. He graduated from University of Zagreb, Faculty of Electrical Engineering and Computing in 2010. From March 2011 to December 2012 he worked as an expert associate on international FP7 project „ANIMPOL - Biotechnological Conversion of Carbon Containing Wastes for Eco-Efficient Production of High-Value Products“ at Faculty of Food Technology and Biotechnology in Zagreb. He is presently employed as a research associate at Faculty of Electrical Engineering and Computing on IPA project “VISTA – Computer Vision Innovations for Safe Traffic”, where he works from April 2013. His research interests include pattern recognition, biometrics, computer vision, as well as biochemical engineering and metabolic network modelling. He is a co-author of several scientific papers in international scientific journals and conferences. He has a full professional proficiency in English, as well as limited working proficiency in German and Hungarian.