Masterarbeit - Explainability for 3D scene understanding

The Fraunhofer-Gesellschaft (www.fraunhofer.de) is one of the world's leading organizations for application-oriented research. 74 institutes develop pioneering technologies for our economy and society—more precisely, 30,000 people from engineering, science, administration, and IT. They know: Anyone who joins Fraunhofer wants to and can make a difference. For themselves, for us, and for the markets of today and tomorrow. At the Fraunhofer Institute for Physical Measurement Techniques IPM, we work every day to make optical technologies usable for specific measurement tasks. At the intersection of science, research, and industry, we develop customized solutions and stand for application-oriented research at the highest level. In the Object and Shape Detection department, we develop mobile measurement systems for surveying infrastructure ranging from roads and railways to complex indoor environments. Our development process covers every step. From design and engineering to system integration and automatic interpretation of measurement data. A cornerstone of our systems is the automated, AI-driven interpretation of 3D point cloud data, which requires high-quality training datasets and, ideally, model interpretability. The primary objective of the thesis is to integrate explainability-driven approaches into the training pipeline. These approaches can be based on 3D uncertainty estimation to identify specific spatial regions or concept bottleneck models to analyze intermediate model representations. These are just examples—the thesis is not strictly limited to these approaches. By applying these methods to different domain-specific benchmarks, the resulting thesis will bridge the gap between model interpretability and dataset quality, providing robust, efficient, and transparent 3D point cloud solutions.