Felix Schmitt
Papers
4
Total Citations
41
H-Index
3
About
Felix Schmitt is a researcher specializing in reinforcement learning (RL), robot navigation, and curriculum learning, with a focus on developing intelligent systems capable of operating effectively in complex, real-world environments. His work addresses some of the most persistent challenges in RL, particularly sparse reward signals and long decision horizons that make training autonomous agents notoriously difficult. Among his notable contributions is his research on hierarchical planning frameworks for robot navigation, which leverages high-level task representations such as rough floor plans to improve learning efficiency — a line of work that has attracted 21 citations in its most-cited form and was formally published as "Hierarchies of Planning and Reinforcement Learning for Robot Navigation" in 2021. Schmitt has also made meaningful strides in curriculum learning, proposing a performance-based start state curriculum framework that systematically improves agent learning by intelligently selecting training start states, earning 14 citations. His earlier work on inverse reinforcement learning for robot navigation, dating to 2016, reflects a longstanding commitment to enabling robots to learn adaptive behaviors in populated environments. Collectively, Schmitt's research contributes practical and theoretically grounded solutions to scalable, efficient robot learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Hierarchies of Planning and Reinforcement Learning for Robot Navigation3 citations · 2021
- 4