In the Curiosity Lab, we are curious about curiosity.
Our goal is to understand, model and create curious agents.
By implementing state-of-the-art machine learning algorithms, we build models of curiosity. With these models, we try to understand and promote human curiosity-driven behavior. We implement the same models on robots and watch them interact and learn about themselves, their environment and other agents.
Our interdisciplinary research integrates computer science, developmental and cognitive psychology, and robotics.
We study, implement, expand and integrate supervised, unsupervised and reinforcement learning algorithms.
We investigate infant, chidren and adult curiosity-driven behavior when interacting with novel information and technology and try to promote curiosity-driven behavior.
We use various commercial robots, such as Lego Mindstorm, NAO and Baxter, to create curious agents that constantly, optimally and actively learn about their environment.