Imitation Learning and Interactive Learning

We are investigating new imitation learning methods with a focus on multi-modal behavior and learning from very small training sets. In addition, we try to incorporate interactive human feedback such as preferences in the learning process.

New Paper @ ICLR 23: Adversarial Imitation Learning with Preferences

In this paper we propose a novel algorithm that extends Adversarial Imitation Learning to use preferences as a feedback besides demonstrations. Results show that our method can learn from expert and imperfect demonstrations. Experiments show the method's effectiveness on robotic ma benchmarks.

CoRL 22: Inferring Versatile Behavior from Demonstrations by Matching Geometric Descriptors

We combine a mixture of movement primitives with a distribution matching objective to learn versatile behaviors that match the expert’s behavior and versatility.

Expected Information Maximization
ICLR 2020: paper accepted - Expected Information Maximization

 Using the I-Projection for Mixture Density Estimation. Find out why maximum likelihood is not well suited for mixture density modelling and why you should use the I-projection instead.