Benedikt Schwab
Papers
1
Total Citations
2
H-Index
1
About
Benedikt Schwab is a researcher at the forefront of computer vision and geoinformatics, specializing in the automated reconstruction of high-detail semantic 3D building models. His work addresses a critical bottleneck in urban modeling: the completion of incomplete facade data. Schwab’s most notable contribution, “FacaDiffy: Inpainting unseen facade parts using diffusion models” (2025), introduces an innovative deep learning approach that leverages diffusion models to plausibly fill in missing or occluded facade elements, such as windows and doors, from 2D conflict maps. This method directly tackles the real-world challenge of sparse or noisy sensor data, enabling more robust and realistic 3D building reconstructions for applications in robotics, autonomous navigation, and urban planning. While his work is still emerging—with his top-cited paper already garnering 2 citations—Schwab’s integration of generative AI into geoinformatics represents a significant step forward. His research bridges the gap between raw, incomplete sensor inputs and the semantically rich, detailed models required for advanced spatial analysis and digital twin creation.
Research Focus
Key Achievements
Top Papers
- 1FacaDiffy: Inpainting unseen facade parts using diffusion models2 citations · 2025