Andrea Vedaldi
Oxford professor who co-leads the Visual Geometry Group; Amazon Scholar; co-author of DINOv3 and VGGT
- Professor of Computer Vision and Machine Learning; co-lead of the Visual Geometry Group, University of Oxford12
- Amazon Scholar, Frontier AI & Robotics (FAR) lab (since September 2026), Amazon1
Andrea Vedaldi is Professor of Computer Vision and Machine Learning and a co-lead of the Visual Geometry Group (VGG) in the Engineering Science department at the University of Oxford, researching computer vision and machine learning methods with a focus on spatial intelligence. In September 2026 Vedaldi started as an Amazon Scholar in Amazon's Frontier AI & Robotics (FAR) lab. Vedaldi is a co-author of Meta's DINOv3 vision foundation model and of VGGT, a feed-forward 3D reconstruction network whose code and models are public. The Royal Academy of Engineering elected Vedaldi a Fellow in 2025; its fellowship page also listed Vedaldi as a research scientist and technical lead at Meta at that time.134562
Work
DINOv3Open weights
Listed among the authors of the DINOv3 paper (2025) and in the repository README, which releases the code and model weights under the DINOv3 License.34
VGGT (Visual Geometry Grounded Transformer) (external site: github.com)Published research
Co-author of "VGGT: Visual Geometry Grounded Transformer" (2025), whose abstract says code and models are publicly available; the repository README lists Vedaldi among the authors.56
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.