I am a postdoctoral research fellow at the Australian Centre for Robotic Vision, at Queensland University of Technologyin Brisbane, Australia. My current research focuses on deep learning techniques for robust robotic vision.

I obtained my PhD degree with Chemnitz University of Technology (chair of Automation Technology, lead by Prof. Peter Protzel) in Germany in April 2012. After two more years as a postdoc in Chemnitz I joined the Robotics and Autonomous Systems Group at QUT in Brisbane in March 2014. Since August 2014 I am working with the Australian Centre for Robotic Vision.

During my PhD thesis I focused on robust graph-based models for SLAM and probabilistic estimation. More recently, place recognition and environmental perception under changing environments has attracted my attention. Within the field of mobile robotics, probabilistic techniques for sensor and data fusion and the problem of simultaneous localization and mapping (SLAM) have been fascinating me for a long time. Related questions of visual perception (using monocular, stereo, or RGB-D cameras), such as structure from motion, place recognition or object recognition interest me as well. The topic of multipath mitigation and how to improve the localization accuracy for satellite-based navigation systems (such as GPS) have been part of my PhD work. Besides all these Bayesian, algebraically well-defined techniques, I find biologically inspired approaches for SLAM and perception very appealing.