Stefan Obwald

Papers

Total Citations

H-Index

About

Stefan Obwald is a roboticist whose research focuses on enabling autonomous mobile robots to intelligently explore and map complex, real-world environments. His work sits at the intersection of autonomous exploration, visual SLAM, and active perception, with a particular emphasis on humanoid and articulated robotic systems. Obwald’s most cited work (82 citations) addresses the challenge of speeding up robot exploration by leveraging background information in the form of topo-metric graphs, a method that is directly applicable to real-world mapping and search tasks. He also introduced DLab, a novel binary descriptor that fuses RGB, depth, and intensity data to achieve robust visual SLAM for humanoid robots, advancing the reliability of localization in challenging conditions. In a third contribution, Obwald tackled the computationally intensive problem of next-best-view coverage for articulated scenes, developing a GPU-accelerated planner that enables robots to manipulate objects to inspect obstructed areas. While his citation counts reflect a focused, early-career impact, his work demonstrates a clear trajectory toward solving practical, high-dimensional problems in autonomous navigation and scene understanding.