Recent Projects
 Oliver works now for Intel, managing project topics including: Rendering, Performance capture, Virtual Production, Machine Learning, Scene Interpretation, Perceptional Rendering, Light-field techniques, VR/AR. For more information see: Intel Visual Computing Institute
Previous Projects
RE@CTThe RE@CT project develops a new paradigm for the production of animated interactive content using multi-camera capture of live action.
PaintingsThe paintings project was a production to deliver animated interactive content using multi-camera capture of live action. This was done together with Queen Mary, University of London.
 Publications: Maleshkova, Julien Pansiot, and Oliver Grau. Through the painter's eye: interactive 3D presentation of paintings on the web. In Proceedings of the 9th Conference on Visual Media Production (CVMP short papers), London, 2012. (PDF)
virtual viewVirtual View generation. Methods for 3D reconstruction and highly realistic rendering of actions in the studio. The methods are based on the fast and robust visual hull computation using the 2D silhouettes of the objects from 12 simultaneously captured images. Current work includes the adaptation of these methods to sport scenes in outdoor environments.
i3DLiveThe goal of the i3DLive project is to develop tools for the multi-camera capture of live action.
iviewThe goal of the iview project is to develop a free-viewpoint system that allows the capture and interactive replay of live events using multiple cameras. The technology uses multi-camera images as an input and is based on algorithms for 3D reconstruction, texture mapping and view-synthesis at interactive refresh rates. Although the techniques could be used for many applications, a sport scenario will be used to demonstrate these techniques.
ORIGAMIIST-ORIGAMI was an EC 5th Framework project which was developing new tools to allow the creation of high-quality scenes incorporating both real and virtual objects. This will enable operations such as lighting changes and camera moves to be carried out in post production, and allows the generation of effects such as the casting of shadows between real and virtual objects. Another aim is to provide visual feedback to actors in a chroma-key environment, so that they can easily relate to the position of virtual objects.
MetaVisionIST-MetaVision was an EC 5th Framework project developing a production system which captures additional image information at the camera to assist in post-production. The BBC's role was primarily concerned with capturing depth information and looking at how it can be applied to help tasks such as integrating real and virtual objects.
digilabDigi-Lab (1997-2000) Integration of virtual objects into image sequences for TV- and film applications
AIDAKnowledge-Based Modelling of 3-D Objects with AIDA (1993-1997)
PANORAMAACT-PANORAMA 1995-1997 : 3-D Telepresence
VRML3-D Information Systems (1995) using VRML
TOFISInteractive Modelling of Buildings (1994-1995)
Plate detectionAutomatic Interpretation of Car Licence Plates (1991-1992)
Fish cellClassification of cytologic specimen with neural networks (1990-1991)