Lourdes Agapito
Also known as Lourdes de Agapito Vicente
Lourdes Agapito is a computer vision professor and Synthesia co-founder whose work spans 3D reconstruction and AI video. After doctoral study in Madrid and a Marie Curie postdoctoral fellowship at Oxford, she joined Queen Mary and led HUMANIS, an ERC project on deformable 3D reconstruction from monocular sequences. At UCL, she is Professor of 3D Vision and led the Vision and Imaging Science Group by 2023. Her later work includes Spann3R, a system that uses spatial memory for dense 3D reconstruction from image collections. She chaired CVPR's 2016 program and, as of October 2026, serves as an ELLIS Program Director and Fellow and a Royal Academy of Engineering Fellow.
- Now at
- University College London, Synthesia and European Laboratory for Learning and Intelligent Systems
- Works
- 158
- Citations
- 6,063
- h-index
- 40
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
By 1997
From Madrid to Oxford
She completed bachelor's, master's and doctoral degrees at the Universidad Complutense de Madrid and joined Oxford's Robotics Research Group as a Marie Curie postdoctoral researcher in 1997.
-
2001
Joining Queen Mary
She joined Queen Mary University's Department of Computer Science in 2001.
-
2008–2014
Reconstructing moving bodies
She led the ERC Starting Grant HUMANIS, which focused on theoretical and practical approaches to deformable 3D reconstruction from monocular sequences.
-
Since 2013
Building vision at UCL
She joined UCL's Department of Computer Science in 2013 and is Professor of 3D Vision there. By 2023, she led the Vision and Imaging Science Group; her research focused on inferring 3D information from images and video.
-
2016
Leading a vision conference
She served as Program Chair for CVPR 2016 and as an area chair for conferences in computer vision.
-
2017
Bringing AI into video
In 2017, she co-founded Synthesia, a London synthetic-media startup that develops AI video and avatar products.
-
2024
Giving scenes spatial memory
With Hengyi Wang, she co-authored Spann3R, a dense 3D reconstruction method that uses external spatial memory to track relevant scene information across image collections.
-
As of October 2026
A role across European AI
ELLIS lists her as a Program Director and Fellow for Learning for Graphics and Vision at UCL. In 2026, the Royal Academy of Engineering elected her a Fellow.
Sources and credits
Photo: robots.ox.ac.uk; Dr Lourdes de Agapito's home page; creator not stated. License: Not stated by source. Source: robots.ox.ac.uk. Cropped to a square and resized. Ask us to remove or replace this photo.