Abstract Lorenzo Jamone 30012017
Cognitive robots that learn: like humans, from humans, with humans
Lorenzo Jamone, Queen Mary University
Humans (and their amazing brains!) can be an inspiration for robotic models of learning and control; also, humans can teach robots, and interact with them. Also, humans can prepare sandwiches (and eat them!). I will outline a few bio-inspired computational models of sensorimotor learning and control that can support intelligent robot behaviors: e.g. reaching, manipulation, tool use and problem solving. The models have been implemented on different humanoid robots using a mix of machine learning, control and multi-modal robot perception (vision, proprioception, touch). Also, I may touch upon a few ideas for human-robot interaction/collaboration and robot learning by human demonstrations.
Lorenzo Jamone recently joined EECS as Lecturer in Robotics, and he is interested in robot cognition, with a focus on sensorimotor learning and perception (body schema, affordances, eye-hand coordination, manipulation, tool use), and in some aspects of human-robot interaction.