Andre Lemme
Papers
13
Total Citations
227
H-Index
9
About
Andre Lemme’s research lies at the intersection of robot learning, dynamical systems, and human-robot interaction, with a focus on enabling robots to acquire complex motor skills from demonstration. His major contributions include developing neural learning schemes for stable dynamical systems, which allow robots to learn and generalize movements while guaranteeing stability—a critical requirement for safe physical interaction. His most cited work, “Neural learning of stable dynamical systems based on data-driven Lyapunov candidates” (52 citations), introduced a method to estimate stable nonlinear dynamics from demonstrations, directly addressing the challenge of ensuring robot motions remain robust and predictable. Lemme also pioneered open-source benchmarking for reaching motion generation (33 citations), providing standardized tools to evaluate and compare learning approaches. His work on self-supervised bootstrapping of movement primitive libraries and neural conditioning for delayed rewards demonstrates his commitment to building autonomous, adaptive systems that learn from sparse, noisy feedback. Through these contributions, Lemme has advanced the practical deployment of learning-based control in robotics, making his research essential reading for those interested in data-driven skill acquisition and stable robot behavior.
Research Focus
Key Achievements
Top Papers
- 1
- 2Neural learning of vector fields for encoding stable dynamical systems45 citations · 2014
- 3Open-source benchmarking for learned reaching motion generation in robotics33 citations · 2015
- 4
- 5
- 6
- 7
- 8
- 9
- 10