THE OXFORD PERCEPTUAL
INTELLIGENCE & EXTENDED
REALITY LAB
The Oxford Perceptual Intelligence and Extended Reality Lab (PIXL) is a computer vision research group in the Department of Computer Science at the University of Oxford.
The team at PIXL has extensive experience developing state-of-the-art algorithms for
perception systems.
Specifically, we work on topics including photorealistic scene capture, vision-language reasoning models, uncertainty, and optimization for vision systems.
Enabling the next generation of machine perception capabilities by combining geometric vision with generative AI.
PIXL is led by Ronald Clark, Associate Professor in the Department of Computer Science at the University of Oxford
OUR RESEARCH FOCUS
-
Realtime Visual SLAM
We are interested in developing algorithms that can simultaneously determine the position of a camera and create a map of the surrounding environment at the same time. This is key for applications such as robotics and autonomous navigation.
-
Vision-Language Reasoning Models
We are developing methods to provide VLMs with rich, three-dimensional information about the world, enabling better spatial understanding and interaction in complex environments.
-
Light, Material and Shape Estimation
Accurately capturing the physical characteristics of real-world scenes is essential for graphics and AR/VR applications. In this direction, we work on estimating lighting conditions, intrinsic material properties, and detailed geometric shapes from images.
-
Physical Science Applications
We are also interested in applying computer vision to scientific challenges, such as reconstructing fluids and clouds from images, opening new ways to study physical processes.
