Title: Foundation Models in Robotics

Speaker: Asst. Prof. Georgios Th. Papadopoulos (Harokopio University of Athens, Greece)

Abstract: The recent introduction of the so-called Foundation Models (FM), i.e. large-scale machine learning models (typically neural network architectures) that have been constructed using vast amounts of (multi-modal) training data and which can be applicable to a wide range of different/diverse tasks, has revolutionized research and technological solution development in the broader field of AI and beyond. The current talk will focus on how FMs are shaping research and development activities in the particular field of robotics. Specifically, the fundamental considerations and key benefits behind the use of FMs in robotics, as well as a brief overview of the recent research in the field, will be initially discussed. Additionally, a detailed discussion regarding the main robotic FM learning methodologies will be provided (namely pre-training, fine-tuning, reinforcement learning and imitation learning), along with the main strengths and limitations of each category of methods. Moreover, key FM-based methods for various robotics tasks (e.g. perception, natural language instructions, reasoning, and control) will be outlined. The talk will conclude with a summary of the current challenges and future research directions in the field.

Short Biography: Georgios Th. Papadopoulos received the M.Eng. and Ph.D. degrees in electrical and computer engineering from the Aristotle University of Thessaloniki, Greece, in 2005 and 2011, respectively. He is currently an Assistant Professor at the Department of Informatics and Telematics of the Harokopio University of Athens, Greece. His research interests include computer vision, machine/deep learning, robotics and artificial intelligence. He has more than 100 publications in international academic conferences and scientific journals. He has over 20 years of experience in participating to EU-funded R&I projects in the areas of ICT, security and robotics. He is a member of the IEEE and the Technical Chamber of Greece