Current page: Members->List of Members->John Gilby
 

John Gilby
Sira Ltd.
Membership Number: 43
Address: South Hill, Chislehurst, Kent BR7 5EH, United Kingdom
Email: john.gilby@sira.co.uk
Phone: +44 20 8468 1714
Fax:
URL: http://www.sira.co.uk

Biographical Sketch
Dr Gilby is a member of the Executive Committee of the BMVA, and the Committees of the UKIVA and the IEE Visual Information Engineering Profession Network. He is also a participant in the Vision Engineering group that is focussed on understanding how robust Computer Vision systems can be built by properly characterising the performance of algorithms.

Sira Ltd.
Sira Ltd is an independent, not-for-profit research and technology organisation (RTO) operating in the area of instrumentation and sensors. We have particular expertise in the application of optically-based instrumentation, including vision systems in applications including industrial inspection, monitoring for security, and tools for the creative media industry. As an RTO our mission is to enable our members and customers to access technology, undertaking work ranging from applied research to prototype design and small volume manufacture. We have capability in both hardware and software design, including embedded and real-time systems.

We understand Cognitive Vision to be Computer Vision operating at the task level. Understanding the scene from a functional point of view - “What does it mean?”

In our work this is instantiated in the recognition and categorisation of objects, structures and events in images and video; learning and adaptation; memory and representation of knowledge; and control and attention. The key idea is to use AI techniques in order to build more robust systems and to try to understand the scene.

Recent work includes:

- The application of AI techniques to build controllers for industrial instruments.

- Use of an agent-based approach to the design of Computer Vision systems.

- Use of uncertainty techniques in industrial inspection.

- Genetic programming.

- View synthesis for object recognition in unstructured 3D environments.

- Intelligent systems for the understanding of industrial data sets.

Some of this work is undertaken for industrial collaborators/customers, but we also carry out more fundamental work in conjunction with University College London, with whom we have co-supervised over 80 PhD students. Key collaborators relevant to this application include Profs Bernard Buxton and John Campbell (Dept of Computer Science).

We have a long track record in the application of AI in engineering situations. For example we developed LINKMAN, an Expert System for the control of cement kilns and other industrial processes, one of the first such products to be successfully commercialised. We also initiated “Design to Inspection” an industry/academia collaboration to develop a system of building knowledge representations that enabled the inspection of engineering parts to be directly linked to their functional use.

Sira is an active participant in vision community groups. Dr Gilby is a member of the Executive Committee of the BMVA, and the Committees of the UKIVA and the IEE Visual Information Engineering Profession Network. He is also a participant in the Vision Engineering group that is focussed on understanding how robust Computer Vision systems can be built by properly characterising the performance of algorithms. Sira is also a managing partner in the Imaging Faraday Partnership, a UK government initiative to bridge the gap between the research base and the commercial exploitation of vision systems. Participants include major UK vision research groups and a diverse range of users of vision technology such as the Police Scientific Development Branch, GlaxoSmithKline, BTxact and Sharp Laboratories of Europe.


Site generated on Friday, 06 January 2006