1 results match your criteria: "Scuola Universitaria Professionale della Svizzera Italiana Manno[Affiliation]"
Curiosity driven reinforcement learning for motion planning on humanoids.
Front Neurorobot
January 2014
Dalle Molle Institute for Artificial Intelligence Lugano, Switzerland ; Facoltà di Scienze Informatiche, Università della Svizzera Italiana Lugano, Switzerland ; Dipartimento Tecnologie Innovative, Scuola Universitaria Professionale della Svizzera Italiana Manno, Switzerland.
Article Synopsis
- Previous research on artificial curiosity (AC) primarily revolves around basic theories and is often limited to simple experimental scenarios.
- The study introduces a curious agent embodied in the complex iCub humanoid robot, utilizing a novel reinforcement learning framework that combines both reactive control and high-level curiosity-driven exploration.
- This is the first instance of an embodied curious agent capable of real-time motion planning, demonstrating intelligent exploration by learning Markov models and showing interest in physical constraints and environmental objects.