Max Robotics — Global Service Robot Directory + US Certification

About

Marcello Pelillo is a leading figure in computer vision and artificial intelligence, best known for pioneering the application of game theory to social scene understanding. His core research focuses on developing robust, mathematically principled methods for detecting and analyzing human interactions in visual data. Pelillo’s most significant contribution is his game-theoretic framework for identifying conversational groups, or F-formations, in images and video sequences. His seminal 2015 paper, "Detecting conversational groups in images and sequences: A robust game-theoretic approach," has garnered 61 citations, establishing a foundational approach for modeling complex social dynamics as evolutionary games. This work, alongside his probabilistic game-theoretic model (51 citations), provides a powerful alternative to traditional clustering methods by naturally handling the ambiguities and spatial constraints of real-world group interactions. More recently, Pelillo has addressed the critical challenge of trustworthy AI, co-authoring "Machines We Trust: Perspectives on Dependable AI" (2021), which explores how to design reliable and ethical autonomous systems. Through his innovative fusion of game theory and computer vision, Pelillo has created a robust toolkit for machines to perceive and interpret the subtle, unspoken rules of human social gatherings.