Benedict Stephan
Papers
9
Total Citations
67
H-Index
4
About
Benedict Stephan is an active researcher at the intersection of robotics, computer vision, and human-robot collaboration, with a particular focus on enabling robots to perceive, understand, and safely interact with humans in real-world environments. His work spans several interconnected areas, including human action recognition, robotic perception, grasp detection, and multimodal sensing. Among his most notable contributions is the ATTACH dataset (2023, 17 citations), a specialized resource for training models to recognize two-handed assembly actions — a critical capability for collaborative robots operating in industrial manufacturing settings. His research on multimodal point cloud segmentation (2021, 16 citations) demonstrated how combining RGB, depth, and thermal data can enable safer human-robot object handovers, a safety-critical challenge in cobot deployment. His PanopticNDT framework (2023, 13 citations) advances panoptic mapping for mobile robots navigating complex indoor environments. Stephan has also contributed to grasp detection, skeleton-based action recognition, and inverse kinematics learning for redundant robots. Collectively accumulating over 65 citations, his body of work reflects a consistent commitment to bridging the gap between theoretical machine learning and the practical demands of autonomous robotic systems in collaborative, human-centered settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3PanopticNDT: Efficient and Robust Panoptic Mapping13 citations · 2023
- 4Fusing Hand and Body Skeletons for Human Action Recognition in Assembly5 citations · 2023
- 5
- 6
- 7
- 8GraspTrack: Object and Grasp Pose Tracking for Arbitrary Objects2 citations · 2024
- 9