Augsburg, Bavaria, Germany
Mentoring students on topics related to perception algorithms: - Evaluation of existing perception algorithms - Machine learning for lane detection - Cone position estimation using a mono camera
Lead programmer of the ROS-based autonomous racing software stack. Developed reliable perception algorithms for cone detection, mapping and lane detection. With our great team we won the prestigious Formula Student Germany 2021 dynamics event. Key areas: - LiDAR Cone Detection - Lane Detection - Fusion of Camera and LiDAR detections - Automatic Camera/LiDAR extrinsic calibration - GraphSLAM
Working Student at Object Recognition Lab.
Bachelor-Thesis on "Large-scale and globally consistent registration of 3D LiDAR point clouds", the resulting algorithm was merged into Blickfeld's production software.
Intern (Praxissemester) LiDAR perception algorithms at Object Recognition lab, supervided by Dr. Hamidreza Houshiar. Did literature research and implemented successfully a point cloud segmentation algorithm tailored to Blickfeld's LiDAR sensor that significantly reduced the amount of noise.
Developed algorithms for mapping LiDAR cones and lane detection based only on the LiDAR-sensor.