Our research revolves around computational methods to convert image-like sensor data into models of the world. The data we are interested in comes from various sources, including consumer cameras, airborne mapping cameras, laser scanners, range cameras, optical and RADAR systems on remote sensing satellites, etc. The output we aim to deliver are maps and models of the environment, such as virtual 3D object models, or maps of elevation and land-cover. We focus on automatic image interpretation, i.e. we aim to create computer systems which can, with minimal user interaction, extract both accurate geometric measurements and useful semantic interpretations of the observed world from the image data. For details, please check the individual project pages, or look up our recent publications.