Matthew Tobias Harris

Queen Mary University of London, City, University of London

Papers

Total Citations

H-Index

About

Matthew Tobias Harris is a pioneering researcher at the intersection of social robotics, human-robot interaction, and computer vision. His work uniquely bridges the technical and the performative, exploring how robots can understand and participate in complex social dynamics. Harris’s most notable contribution is the creation of the **ORBIT dataset** (2021, 32+ citations), a real-world, few-shot benchmark for teachable object recognition. This dataset directly addresses a critical bottleneck in robotics and personalization: enabling machines to learn new objects from just a handful of examples, moving beyond the data-hungry paradigms of traditional deep learning. Complementing this technical achievement, his earlier work on the **Robot Comedy Lab** (2015, 55 citations) remains his most cited. In this study, Harris used live comedy performance as a testbed to experimentally probe the subtle, coordinated social signals—gaze, gesture, body orientation—that underpin successful audience interaction. By deploying robots in this high-stakes social context, he demonstrated a novel methodology for studying non-verbal communication. Harris’s research is distinguished by its creative, interdisciplinary approach, using real-world scenarios to push the boundaries of both social robotics and machine perception.