A new training centre will embed PhD students inside companies that build video games, medical scanners, and augmented reality systems, rather than keeping them in university labs. The problem is that visual engineering—using computers to interpret images, generate graphics, or model complex designs—demands skills from multiple disciplines. A standard PhD rarely provides that breadth. Students may learn computer vision but not how to integrate it with real-time graphics, or understand 3D modelling but not how to apply it to pedestrian flow in a building. The centre addresses this gap by structuring training around industry partnerships, so students work on concrete problems—object detection for food sorting, body measurement from scans, or visualisation of geographic data for urban planning—while drawing on expertise from across UCL’s imaging, interaction, and computational science groups. If successful, the centre will produce engineers who can move seamlessly between research and application. That could accelerate development of augmented reality tools for defence, heritage, and entertainment; improve medical image interpretation; and make architectural design more responsive to data. The immediate impact is on the people trained, not on a single product.
View original technical description
In redeveloping the EngD VEIV centre, we will be focussing on three themes in the area: - Vision & Imaging, covering the areas of computer-based interpretation of images. For example, object tracking in real-time video, or face detection and surface appearance capture. UCL now has a broad expertise in medical imaging (see description of CMIC), and also in tracking and interpretation of images (e.g. expertise of Julier and Prince who are on the management team). Previously we have supported several EngD projects in this area: e.g. Philips (structure from MRI), Sortex (object detection), Bodymetrics (body measurement from scanning data), where the innovation has been in higher-levels of interpretation of imaging data and derivation of measurements automatically. Two other projects highlight the rapidly developing imaging technology, with high-density sensors and high dynamic range imagery (e.g. BBC and Framestore). We have outline support from several companies for continuing in this area. - Media & Interfaces, covering real-time graphics and interactive interfaces. For example, the use of spatially immersive interfaces, or computer games technology. We have a growing relationship with a number of key games companies (EA, Sony, Eidos, Rebellion), where their concern or interest lies in the management of large sets of assets for complex games software. There is interest in tools for developing imagery (r.g. Arthropics, Geomerics). We also have interest in the online 3D social spaces from IBM and BT. A relatively recent development that we plan to exploit is the combination of real-time tracking, real-time graphics and ubiquitous sensing to create augmented reality systems. Interest has been expressed in this area from Selex and BAe. There is also a growing use of these technologies in the digital heritage area, which we have expertise in and want to expand. - Visualisation & Design, covering the generation and visualisation of computer models in support of decision-making processes. For example, the use of visualisation of geographic models, or generative modelling for architectural design. Great advances have been made in this area recently, with the popularity of online GIS tools such as Google Earth tied in to web services and the acceptance of the role of IT in complex design processes. We would highlight the areas of parameterised geometry (e.g. with Fosters and the ComplexMatters spin-out), studying pedestrian movements (with Buro Happold, Node Architects), visualisation of GIS data (e.g. ThinkLondon, Arup Geotechnical), and medical visualisation.These themes will be supported by broadening the engagement with other centres around UCL, including: the UCL Interaction Centre, the Centre for Medical Image Computing, the Chorley Institute and the Centre for Computational Science.The main value of the centre is that visual engineering requires cross-disciplinary training. This is possible with a normal PhD, but within the centre model inter-disciplinary training can embed the students' focussed research into a larger context. The centre model provides a programme structure and forums to ensure that opportunities and mechanisms for cross-disciplinary working are available. The centre also provides an essential role in providing some core training; though by its nature the programme must incorporate modules of teaching from a wide variety of departments that would otherwise be difficult to justify.
Alan Penn (Co-Investigator)Alasdair Turner (Co-Investigator)Anthony Steed (Principal Investigator)Simon Julier (Co-Investigator)Simon Prince (Co-Investigator)
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know