Georg Martius received his Diploma (Master equivalent) in Computer Science from University of Leipzig, Germany, in 2005. He pursued a Ph.D at the Bernstein Center for Computational Neuroscience in Göttingen, Germany finishing in 2009 with highest praise. Several postdoc positions at the Max Planck Institute (MPI) for Dynamics and Self-Organization, Göttingen, at the MPI for Mathematics in the Sciences in Leipzig and at the Institute of Science and Technology Austria have broadened his background. Since 2017 he is leading the Research group in Autonomous learning at MPI for Intelligent Systems in Tübingen, Germany, with the focus on Machine Learning for robotics.
Education and positions held
- 2005 – 2009
- PhD at Bernstein Center for Computational Neuroscience (BCCN) Göttingen, Germany.
- 1999 – 2005
- Diploma in Computer Science.University of Leipzig, Faculty of Mathematics and Computer Science.
- 2002 – 2003
- Visiting Student, University of Edinburgh, Division of Informatics, Scotland, UK.
- 2017 – present
- Max Planck Research Group Leader, MPI for Intelligent Systems, Tübingen.
- 2015 – 2017
- IST-Fellow at Institute of Science and Technology Austria.
- 2010 – 2015
- Post-doctoral position at Max Planck Institute for Mathematics in the Sciences, Leipzig, Germany.
- 2009 – 2010
- Post-doctoral position at Max Planck Institute for Dynamics and Self-Organization, Göttingen, Germany.
I am interested in autonomous learning, that is how an embodied agent can determine what to learn, how to learn, and how to judge the learning success. I am using information theory and dynamical systems theory to formulate generic intrinsic motivations that lead to coherent behavior exploration – much like playful behavior. I am working on machine learning methods particularly suitable for reinforcement learning, internal models, and representation learning.
- G. Martius, R. Der, and N. Ay. Information driven self-organization of complex robotic behaviors. PLoS ONE, 8(5):e63400, 2013.
- R. Der and G. Martius. Novel plasticity rule can explain the development of sensorimotor intelligence, PNAS, 2015, doi: 10.1073/pnas.1508400112
- S. S. Sahoo, C. H. Lampert, and G. Martius. Learning equations for extrapolation and control. In Proc. Intl. Conf. on Machine Learning (ICML’18), volume 80, pages 4442-4450, PMLR, 2018
- M. Rolinek, D. Zietlow and G. Martius. Variational Autoencoders Recover PCA Directions (by Accident), In Proc. Conference on Computer Vision and Pattern Recognition (CVPR’19), pages 12406–12415, 2019
- S. Blaes, M. Vlastelica-Pogančić, J.-J. Zhu, and G. Martius. Control What You Can: Intrinsically motivated task-planning agent. In Advances in Neural Information Processing Systems 32 (NeurIPS’19), 2019.