Benjamin Busam
Technical University of Munich, Munich Center for Machine Learning
Papers
29
Total Citations
449
H-Index
15
About
Benjamin Busam is a researcher whose work spans computer vision, robotics, and medical imaging, with particular expertise in 6D object pose estimation, robotic grasping, and surgical autonomy. His contributions to category-level pose estimation have been especially influential: datasets like PhoCaL (45 citations) and HouseCat6D (26 citations) have provided the research community with essential benchmarks for evaluating pose and shape estimation on photometrically challenging and household objects. His work on CPS++ (35 citations) advanced monocular class-level pose estimation through self-supervised learning, while MonoGraspNet (42 citations) demonstrated that robust 6-DoF grasping is achievable from a single RGB image. Beyond manipulation, Busam has made notable contributions to few-shot robotic learning through DemoGrasp (29 citations) and scene understanding via the scene-graph-driven rearrangement framework SG-Bot (24 citations). His research also extends into medical robotics, including autonomous systems for retinal surgery, robotic thyroid volumetry using ultrasound, and gamma-imaging-guided needle biopsy. This breadth reflects a unifying ambition: developing intelligent, perception-driven robotic systems that operate reliably across both industrial and clinical environments, making him a distinctive voice bridging fundamental computer vision research and real-world robotic deployment.
Research Focus
Key Achievements
Top Papers
- 1
- 2MonoGraspNet: 6-DoF Grasping with a Single RGB Image42 citations · 2023
- 3
- 4DemoGrasp: Few-Shot Learning for Robotic Grasping with Human Demonstration29 citations · 2021
- 5RSV: Robotic Sonography for Thyroid Volumetry29 citations · 2022
- 6
- 7
- 8
- 9Cooperative Robotic Gamma Imaging: Enhancing US-guided Needle Biopsy20 citations · 2015
- 10ColibriDoc: an Eye-in-Hand Autonomous Trocar Docking System18 citations · 2022