Doctoral colloquium - Radovan Gregor (9.11.2026)
Monday 9.11.2026 at 13:10, Lecture room I 9
Radovan Gregor:
Mirroring mechanisms for trustworthy human-robot interaction
Abstract:
The discovery of neurons in the premotor cortex of macaque monkeys that activates both when the animal performs a goal-directed action and when it observes another individual performing the same action suggests that the brain contains a system that “mirrors” the actions of others. This system became known as the Mirror Neuron System (MNS) (Rizzolatti G., 1992). The finding proposes that action perception and action execution share a common neural representation. Understanding what another agent is doing may then rest on the observer’s own motor repertoire, not on a separate visual analysis. The idea has strongly influenced our view of the neural basis of social cognition, imitation and intention understanding.
Inspired by the MNS, we propose an architecture for motor-primitive discovery and action recognition on the NICO humanoid robot. The central premise is that the robot should interpret an observed movement through the same compact motor representation it uses to produce movement.