Papers
22
Total Citations
320
H-Index
10
About
Hermann Blum is a versatile robotics researcher whose work spans neuromorphic computing, robotic perception, and construction automation. Early in his career, Blum made significant contributions to neuromorphic systems, demonstrating how brain-inspired analog/digital hardware could enable low-power, event-driven robot navigation — work that garnered over 90 citations and established him as a notable voice in the field. His research showed that neuromorphic controllers equipped with dynamic vision sensors could achieve robust obstacle avoidance and target acquisition while dramatically reducing energy consumption compared to conventional architectures. Blum subsequently expanded his focus to semantic scene understanding, tackling critical challenges in out-of-distribution detection, sensor fusion, and domain adaptation. His embodied active domain adaptation framework, which autonomously adapts segmentation networks to unseen environments through intelligent path planning, reflects his broader interest in making robots reliably self-sufficient in real-world settings. More recently, Blum has applied his expertise to construction robotics, developing a high-accuracy structured light sensor and building model synchronization tools that bring millimeter-level precision to on-site automation. His mixed reality human-robot teaming interface further highlights his commitment to practical deployment. With over 270 cumulative citations across diverse domains, Blum's research consistently bridges cutting-edge algorithms with real-world robotic applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5A 3D Mixed Reality Interface for Human-Robot Teaming22 citations · 2024
- 6
- 7Modular Sensor Fusion for Semantic Segmentation14 citations · 2018
- 8
- 9
- 10See Yourself in Others: Attending Multiple Tasks for Own Failure Detection11 citations · 2022