Bastian Bischoff

Papers

Total Citations

H-Index

About

Bastian Bischoff is a robotics researcher focused on advancing autonomous robot navigation and control through machine learning. His work bridges computer vision, reinforcement learning, and probabilistic modeling to enable robots to operate reliably in complex, uncertain environments. A key contribution is his development of a sensor fusion approach that combines vision and odometry for accurate indoor robot localization, demonstrating that cost-effective systems can achieve high precision without expensive laser scanners or motion-capture hardware (23 citations). Bischoff has also made significant strides in robot learning, introducing a policy search method that efficiently learns control policies from sparse data—critical for real-world applications like grasping and manipulation where data is limited (20 citations). His hierarchical reinforcement learning framework for robot navigation addresses the curse of dimensionality, making complex tasks tractable by decomposing them into manageable sub-problems (13 citations). Additionally, Bischoff pioneered Probabilistic Value-Iteration (PVI), a novel approach that uses Gaussian Processes to handle continuous state-action spaces under uncertainty, advancing the theoretical foundations of reinforcement learning. His work has practical implications for service robotics, where affordable, robust autonomy is essential.