Human-Computer Interaction (A5.1)
Robot interaction (A5.10.5)
Machine Learning (A9.2)
The synergy between artificial intelligence and robotics allows the development of robots with increased autonomy and efficiency. Relevant applications require that algorithms are embedded and operate in real time while using partial sensory data of reduced quality. Complementary sensors and actuators are brought by ambient computing, home automation systems and connected objects. In this context, our main activities within Lab-STICC are related to ambient devices interconnection, automatic discovering and versability management, to environment mapping via computer vision, robot learning by demonstration or imitation and algorithm optimization for faster processing, with a particular focus on the use of robots for assisting humans in their daily tasks. We study how ambient devices and robots can build increasingly complex skills based on active learning, artificial curiosity and interactive learning, to allow continuous lifelong learning of robots in open and connected environments and in interaction with humans.