CV

University of Bonn

Ph. D. Candidate

Computer Science

• Supervisor: Prof. Dr. Juergen Gall
• Topic: Understanding Traffic Scenes by Means of Graph-Networks
• Research interests: Graph Neural Networks, 3D Multi-Object Tracking, End-to-End Autonomous Driving
• Expected graduation: April 2025

Karlsruhe Institute of Technology (KIT)

Master of Science (M. Sc.)

Mechatronics and Information Technology

• Minor: Robotics
• Master Thesis: Development of Single Stage Detectors on Multi-Layer Grid Maps
• Average grade: 1.5 (scale: 1.0 - excellent, 4.0 - sufficient)

Karlsruhe Institute of Technology (KIT)

Bachelor of Science (B. Sc.)

Mechanical Engineering

• Minor: Mechatronics
• Bachelor Thesis: Evaluation of Algorithms for Registration of Point Clouds
• Average grade: 1.9 (scale: 1.0 - excellent, 4.0 - sufficient)

Mercedes-Benz Group AG

Industrial Doctoral Researcher

• Conducted diverse research initiatives within the realm of autonomous driving, exploring various innovative research topics such as learning Non-Maximum Suppression (NMS) with Graph Neural Networks, learning data association using Graph-Transformers for 3D Multi-Object Tracking (MOT), end-to-end multi-camera object detection and tracking, and end-to-end Bird's Eye View (BEV) future instance prediction.
• Engineered deployment infrastructures for training and evaluation processes on several cloud platforms, including Microsoft Azure, Google Cloud Platform and an in-house GPU cluster, ensuring scalable model deployment.
• Provided academic guidance and mentorship to graduate students, including supervision of one Bachelor's and three Master's Theses along with their associated pre-thesis internships.
• Published and presented research papers at premier Computer Vision and Artifical Intelligence conferences such as CVPR, ICCV and IJCAI, and contributed as reviewer at leading venues.

Forschungszentrum Informatik (FZI)

Graduate Research Assistant

• Improved the results from my master thesis and enhanced the 3D object detection performance on multi-layer grid maps. Developed an innovative unsupervised domain adaptation method based on this refined detector, achieving sensor-agnostic detection capabilities and domain transferability across various datasets, e.g. nuScenes and KITTI.
• Published and presented the research findings in an international peer-reviewed conference.
• Upgraded the existing codebase for several deep learning applications on multi-layer grid maps, migrating from TensorFlow Object Detection API with TensorFlow 1.0 to a custom framework with TensorFlow 2.0.

Forschungszentrum Informatik (FZI)

Graduate Research Assistant

• Designed and constructed mounting components to connect a humanoid robot with a wheeled robot platform using PTC Creo and tested the movement of the integrated robot using ROS Unified Robot Description Format (URDF).
• Conducted reachability analysis using ROS Reuleaux tools to evaluate the robot's operational capabilities.
• Participated as a member of team AlpaKa in the Audi Autonomous Driving Cup 2018 and secured the first place.

SEW-EURODRIVE GmbH & Co. KG

Intern for Product Line Management

• Developed a CAD library of adjustable standard components for factory planning using SketchUp and prepared training documents to facilitate library usage. Constructed workplaces for servomotor assembly using the library.
• Optimized the factory layout considering space requirements, material flow, assembly processes and economic factors.
• Standardized the visualizations of servomotor components with varying sizes and features and configured highlighted displays for special features in SAP Manufacturing Execution (SAP ME), simplifying the identification of components.

Journal Reviewer

• IEEE Sensors Journal

Conference Reviewer

• CVPR 2023, 2024, 2025
• ECCV 2022, 2024
• ICRA 2024
• ACCV 2024
• ICCV 2023