Robots Learning Through Interaction

Explore the cutting-edge field of robot learning through interaction in this 40-minute talk by Prof. Jens Kober from TU Delft, part of the RWTH Artificial Intelligence Colloquium series. Delve into the complexities of robotic skill acquisition, including environmental interactions, high-dimensional data processing, non-linear dynamics, and human-robot collaboration. Discover various methods in supervised learning, imitation learning, reinforcement learning, and interactive learning, illustrated through real-world robot experiments ranging from playful tasks to practical retail applications. Gain insights from Kober's award-winning research on motor skill learning, deep reinforcement learning, and machine learning for control, and understand how human-robot interaction can enhance the efficiency and effectiveness of the learning process.