Leveraging human inputs in interactive machine learning for human robot interaction

Senft, Emmanuel; Lemaignan, S�verin; Baxter, Paul E.; Belpaeme, Tony

Abstract

A key challenge of HRI is allowing robots to be adaptable, especially as robots are expected to penetrate society at large and to interact in unexpected environments with non-technical users. One way of providing this adaptability is to use Interactive Machine Learning, i.e. having a human supervisor included in the learning process who can steer the action selection and the learning in the desired direction. We ran a study exploring how people use numeric rewards to evaluate a robot's behaviour and guide its learning. From the results we derive a number of challenges when designing learning robots: what kind of input should the human provide? How should the robot communicate its state or its intention? And how can the teaching process by made easier for human supervisors?

• Presentation Conference Type: Conference Paper (published)
• Conference Name: ACM/IEEE International Conference on Human-Robot Interaction
• Start Date: Mar 13, 2017
• Acceptance Date: Nov 30, 2016
• Publication Date: Mar 31, 2017
• Deposit Date: Sep 2, 2020
• Pages: 281-282
• Book Title: Proceedings of the 2017 ACM/IEEE Human-Robot Interaction Conference
• ISBN: 9781450348850
• DOI: https://doi.org/10.1145/3029798.3038385
• Public URL: https://uwe-repository.worktribe.com/output/4743034