Patrick Tutzauer

• Working on an approach for large-scale building use classification with Deep Learning • Investigating human understanding of buildings and working on perception-guided adaptive modeling of 3D virtual cities • Developed an algorithm for semi-automatic building reconstruction from point clouds and coarse building models • Teaching Assistant: - Pattern Recognition and Image Understanding - Modelling and Visualization of Geodata

Deep Learning for Urban and Rural Data Interpretation • Working on a Deep Learning approach to estimate hiking speeds based on remote sensing imagery

Internship and Diploma Thesis in the department of stereo camera based advanced driver assistance systems During the internship I developed algorithms, to • Segment collision-relevant areas in front of the car • Measure heights for e.g. parking garage entries, to ensure accessibility Diploma Thesis: • Developed an approach to measure the road surface in front of the car in 3D, detect and warn the driver about imperfections like potholes or speed bumps ahead

Photogrammetry Computer Vision