Computer Vision PhD Candidate
I create synthetic 3D data to train the next generation of foundation models.
I'm a PhD candidate in the Visual Geometry Group at the University of Oxford (supervised by Andrea Vedaldi, Christian Rupprecht, and Iro Laina). During my PhD, I've interned at Meta Reality Labs (supervised by Adam Harley and Jakob Engel).
Under review
A pipeline that generates people carrying arbitrary objects in 3D space by distilling motion from video models. We use it to produce a densely annotated 2.2M-frame dataset from six synchronised egocentric and exocentric cameras.
ECCV 2026
Visually and semantically coherent 3D worlds through scene-scale diffusion with fine-grained layout control. We fine-tune a pre-trained diffusion model to apply it as a convolutional operator for scene-scale 3D generation.
ICCV 2025
Fully navigable 3D worlds beyond the small "bubble" worlds generated by prior methods. We prompt pre-trained 2D image and image-to-3D generators to create immersive environments without any training.
3DV 2025
Geometrically consistent scene generation through depth inpainting. We utilise depth information in the generation process, simplifying previous pipelines while creating more coherent 3D scenes.
NeurIPS 2023 Workshop: Self-Supervised Learning - Theory and Practice
Utilising features learned by self-supervised masked autoencoders (MAE) for unsupervised instance segmentation.
* denotes equal contribution