George Vogiatzis is a Reader in Computer Science at Loughborough University. His research is in computer vision and deep learning, with a long-running focus on recovering three-dimensional structure from images and, increasingly, on learning-based methods for perception, decision-making and multimodal understanding. Over his career this work has spanned multi-view and photometric stereo, self-supervised monocular depth estimation, reinforcement learning for intelligent transport, vision-and-language and visual retrieval, and generative models.
He studied Mathematics and Computer Science at Imperial College London, graduating with a first-class MSci, before completing his PhD in computer vision at Trinity College, University of Cambridge, on the visual estimation of shape, reflectance and illumination. His doctoral studies were supported by a Gates Cambridge Scholarship, and he subsequently held a Junior Research Fellowship at Wolfson College, Cambridge.
From 2006 to 2009 he was a Senior Research Scientist at Toshiba Research Europe in Cambridge, where he was part of the Computer Vision Group working on cultural-heritage digitisation, photorealistic 3D animation and automotive applications. He joined Aston University as a Senior Lecturer in 2010 and moved to Loughborough University as a Reader in 2023.
A consistent theme of his work is its translation into practice. He has led around £2.4M of research funding and collaborated extensively with industry — including Toshiba, Activision/Blizzard and Canon, alongside a large portfolio of Innovate UK Knowledge Transfer Partnerships — on problems ranging from 3D reconstruction and tyre inspection to autonomous-vehicle perception, large-scale image and video retrieval, and air-quality modelling for transport planning. He has published more than 70 papers in leading computer-vision venues such as TPAMI, IJCV, CVPR, ICCV and ECCV, and is a named inventor on five international patents.
He has supervised many doctoral researchers across computer vision, reinforcement learning and multimodal learning, and is a Fellow of the Higher Education Academy. He has taught modules in deep learning, multimedia information retrieval, and multi-agent systems, and developed an MSc in Applied AI aimed at graduates from outside computer science, as well as one of the earlier postgraduate modules dedicated to deep learning in the UK.