The GV2 group in Trinity College is hosting the Cell Centre of Competence in Europe. This project is funded by Enterprise Ireland with STI (Sony-Toshiba-Ibm) . It aims at developing applications for computer games on the Ibm CELL. In particular, we are interested in the realtime robust inference of body pose using one or more cameras. In addition to the visual observations,  additional prior information from motion capture database will also be available for bayesian  inference.

The project aims to develop a Virtual Reality (VR) toolkit which will allow an instantaneous and rich representation of abnormalities in the heart derived from standard electrocardiographic signals.

This project is focussed on exploring the use of Dual Quaternions in computer graphics. Dual Quaternions are an extension to the Quaternion algebra that permits the representation of screw motions (motions with rotation and translation) and thus are a natural means of representing frame blending operations. The project has shown applications of Dual Quaternions for real-time character skinning.

Funded by the Higher Education Authority of Ireland's Programme for Research in Third Level Institutions (HEA-PRTLI), this multidisciplinary collaborative project at Trinity facilitates research in a wide range of different sub-projects. Our main areas of research within this project are Perceptually Adaptive Graphics, Human Animation, Large Scale Urban Simulation and Scientific Visualization.

Geopostors is the Crowd Rendering technology developed by Simon Dobbyn during his PhD research and now we are building on this work to develop a cost-effective tool that could be deployed by games developers to add large realistic crowds to their simulations. This work is funded by a Proof of Concept grant from Enterprise Ireland.

This project addresses the challenge of realistic simulation of crowds of humans. An important factor is that realistic gesture and interactions between the individuals in the crowd are lacking. What needs to be done is to incorporate conversational behaviour into the scene, where individuals communicate with each other in both verbal and non-verbal ways.

The goal of this project is to create an immersive virtual urban environment which is capable of running in real-time on commodity hardware.

This project addressed the use of time-critical, adaptive techniques in Real-time animation applications, such as Virtual Reality (VR) and simulation. In large-scale virtual environments, rendering many objects, generating realistic motion, and handling interactions, imposes an incredibly high computational load. The specific focus is on improving collision detection and response.

This project is funded by Enterprise Ireland and PixAlert. It aims at automatically detecting pornographic content in digital video streams for automatic filtering.

This project took the notion of evaluating motion graphs and expanded it significantly, allowing environments of practical size and complexity with motions and evaluations extending well beyond just locomotion.

This project took the notion of optimizing motion graphs and expanded it significantly, allowing environments of practical size and complexity with motions and evaluations extending well beyond just locomotion.

FINGRID stands for Financial GriID.

This project uses overhead cameras to monitor blackjack play in a casino in order to both monitor play (game statistics) and to detect cheating.

This project seeks to address the problem by using research from cognitive psychology and engineering to model aspects of human vision, memory and attention for the purpose of providing internally driven, autonomous attention behaviours. Behaviours driven in such a way serve both aesthetic and functional needs, forming the basis of a more general framework envisaged for use with a cognitive modelling paradigm.

This project involves implementing elements of a physics engine on a Field-Programmable Gate Array (FPGA), to accelerate physics computation above what is achievable on a commodity CPU. Ultimately, the physics engine will be expanded by distributing it across a cluster of FPGAs, enabling high-speed large-scale physics computation.

This project seeks to automatically index personal digital picture collections with the names of the persons appearing in it. It is involving processes such as face detection, face recognition with continuous interactive learning with the users.