Summary
Gregor Kobsik is a PhD student and researcher specializing in Geometry and 3D Deep Learning at RWTH Aachen, with nine years of experience spanning academic research, teaching, and applied AI training. His work focuses on deep shape representation—shape analysis, modeling, and reconstruction—using unsupervised and self-supervised methods to decompose partial symmetry in 3D geometry, resulting in publications such as a 2024 arXiv paper on contrastive learning for geodesic point cloud patches and a CVPRW 2023 paper on octree transformers. He also teaches practical AI skills through industry-oriented courses on prompt engineering and process automation, and has hands-on experience building VR applications and embedded computer-vision tools. Comfortable with Python and PyTorch as well as C++/Vulkan for systems work, he bridges theoretical research and implementable solutions that support student supervision and industrial projects. Actively seeking a research-centered internship for Fall 2024, he brings both deep domain expertise and a track record of turning complex geometric problems into reproducible models and tools.
9 years of coding experience
Informatik, Informatik at University of Koblenz and Landau
Master Informatik, Master Informatik at RWTH Aachen University
Abitur, Abitur at Rhein-Wied-Gymnasium Neuwied
German, English, Polish