Danica Kragic

Research explores how agents understand their environment via interaction and action.

MetricRL enhances learning from past experiences in goal-oriented tasks.

HyperSteiner presents a more efficient method for calculating Steiner Minimal Trees.

This article examines how relative representations improve AI communication and task adaptability.

New insights reveal the active contribution of visual encoders in robotic decision-making.

Researchers test if transformer models can predict human perceptions of odors.

Learn how Robot-Centric Pooling improves robots' understanding of themselves and their surroundings.