What would you like to talk about? Robotics, Human-Robot Interaction, Human-in-the-loop Reinforcement Learning

Bio: With my background in computer science and psychology, I develop techniques to automatically synthesize corrections for robotic systems. When a robot fails, it can autonomously ask for help and learn how to correct its behavior so that it can continue with its task. This approach requires the robot to interact with people and integrate their feedback into its decision-making on the fly. By endowing robots with the ability to automatically correct and adapt their behavior during their operation through human feedback, my work contributes towards a future of robust, verifiable robots that can help people with daily tasks.

Sanne has previously spoken at a WiDS AI & ML Sweden event.