Jan Fabian Schmid

Universität Hamburg, Robert Bosch (Germany)

Papers

Total Citations

H-Index

About

Jan Fabian Schmid is a robotics researcher whose work bridges the critical gap between autonomous perception and precise localization. His primary research areas include active visual object search, deep reinforcement learning for robotics, and ground texture-based localization. Schmid’s most influential contribution is his deep reinforcement learning framework for active visual object search, which teaches service robots to intelligently explore environments, approach target objects, and terminate tasks—a fundamental capability for real-world deployment. This work has garnered 14 citations and represents a significant step toward making robots more autonomous and task-aware. In parallel, Schmid has advanced ground texture-based localization, developing a self-contained method using compact binary descriptors that enables both global and local localization for vehicles, achieving 11 citations. His most recent work applies deep metric learning to ground images, pushing toward low-cost, high-accuracy self-localization solutions. By combining reinforcement learning for active perception with robust localization techniques, Schmid is helping to create robots that can both understand their surroundings and precisely know where they are—two essential pillars for truly autonomous systems.