8 December 2005

Researchers seek artists' "Holy Grail"

A symphony of sounds which can be raised by the movements of a conductor's hands - but without any orchestra in front of him. An exhibition of images which constantly change depending purely on how many people are looking at them and where they are standing. Art which can be influenced by the response of an audience as much as by the artist's own actions.

Researchers at the University of Dundee have embarked on a project to find a "Holy Grail" for artists, which could make all of the above readily accessible.

They are aiming to create a new gesture-tracking interface for artists and performers - an application that could revolutionise how both artists and their audiences can interact with artifacts, installations or performances.

The G-Vision project is a collaboration between Nigel Johnson, Reader in Interactive Media in the School of Television and Imaging, and Dr Stephen McKenna, of the Division of Applied Computing. It has been supported by a £40,000 Digital Media Commercialisation Award, funded jointly by Scottish Enterprise (SE) Tayside and the University’s own Department of Research and Innovation Services.

The one-year project aims to create a prototype software `plug-in’ which could be used to track anything from a simple hand gesture to the movements of groups of people. The software could then process these movements and respond intelligently in whichever way the artist wants it to.

For instance, specific hand gestures from either an artist working within a real-time installation or giving a performance could trigger many different actions and reactions, tailored to the artist’s requirements, without having to physically handle any equipment.

Similarly, the movements of a large group of people viewing an exhibition could be used to instantly change what they are hearing or seeing.

"This sort of low-cost and effective, real-time, programmable interface has long been a kind of Holy Grail in the interactive arts," said Nigel Johnson.

"This is something which would only be restricted by the artist’s own imagination and how they wish to apply it."

G-Vision would plug-in to software which is currently available, but would advance the facility available to artists.

"There is software available to artists at present which attempts to do some of this, but it is not as sophisticated as what we are trying to develop," said Dr McKenna.

"In this project we are developing technology specifically for artists, and we will be working closely with performers and artists. However, it may in the long-term be shown to have benefits in other areas."

The project has already attracted international attention.

The funding award was one of four Commercialisation Awards made recently funded by SE Tayside and RIS. The award provides funding for technologies with strong commercial potential , so they can be developed it to a stage where they are more attractive for further development. This is the first Commercialisation Award to be made in the area of Digital Media.

Jill Farrell, SE Tayside director business growth, commented: "This project is another fine example of the breadth of expertise to be found in Tayside’s burgeoning digital media sector. We welcome the opportunity to support innovative projects which can be turned into commercially successful products, as these give our companies greater potential to grow and make a greater contribution to Tayside’s economic prosperity."

Nigel Johnson, School of Television & Imaging (Visual Research Centre)
Faculty of Duncan of Jordanstone. n.m.johnson@dundee.ac.uk

Nigel Johnson is an artist and researcher whose work has centered on the production of computer-controlled, two and three dimensional, real-time interactive installations and digital artworks since 1978 and has exhibited his work widely, both nationally and internationally. Appointed Reader in Interactive Media in 2000, based at the Visual Research Centre, he has previous experience of working on large-scale arts-based projects involving the co-ordination of designers, software and hardware engineers as well as the programming, development and production of large and small-scale interactive installations.

Stephen McKenna, Applied Computing

Faculty of Engineering & Physical Sciences.stephen@computing@dundee.ac.uk

Stephen McKenna has been researching the use of computer vision for automated analysis and machine understanding of human activity for a decade. A set of state-of the-art algorithms has been developed through his previous research projects that are suited to real-time interaction via vision-based recognition. Schemes for representing visual activity combined with appropriate filtering and recognition algorithms can enable real-time recognition of various activities. Automatically extracted, time-varying parameters that represent artists' activities in rich and expressive ways can thus be used to drive interactive art that is responsive and natural.

For more information contact:

Roddy Isles,
Head of Press
Tel: 01382 344910,
Email: r.isles@dundee.ac.uk