At the VCG we are always looking to welcome new PhD students. To apply for a PhD at the University of Leeds you generally need to have four things:
Recent work has generalised isosurfaces for scalar field visualisation to fiber surfaces for bi-field visualisation in volumetric data. However, much data, especially from medical acquisition, is commonly analysed in image form. The goal of this PhD is therefore to apply the 2D analogue of fiber surfaces to image analysis, and to compare it with existing methodologies for image-based boundary detection. Contact: Hamish Carr.
For analysis of multiple fields over a single spatial domain, one set of techniques is based on collapsing the individual fields into a single summary field to which scalar field visualisation can be applied. Since the recent introduction of fiber surfaces for bifield visualisation, it becomes feasible to reduce to two fields, not one, thereby preserving more variation in the data. The goal of this PhD is therefore to use this line of attack to deliver further insight into complex high-dimensional simulations. Contact: Hamish Carr.
Ultra-high resolution displays can be desktop-based or wall-sized, and show orders of magnitude more information than is possible with conventional displays. Such displays are a key weapon for the analysis of gigantic datasets, which are now commonplace in all aspects of science and business. However, to be effective we need to integrate novel visualizations on these displays with underlying analysis algorithms. The goal of this PhD is to develop such visual analytic interfaces, using a particular application domain as a case study. Contact: Dr Roy Ruddle.
High- and low-throughput next generation sequencing can now be performed for modest cost, making it feasible to sample hundreds of individuals to understand the causes and development of life-threatening diseases such as cancer. Today it is the scientists themselves rather than sequencing technology that is the bottleneck in the analysis of such sequencing data, due partly to the lack of suitable visualization tools for comparative genomics. The goal of this PhD is to exploit the capability of ultra-high resolution displays to develop such tools. Contact: Dr Roy Ruddle.
There is a history of growth-based procedural modeling for a wide range of geometry domain – from trees to urban environments. Growth based modeling has the advantage of emergent behaviour which is easily understood, and geometrically smooth. Such smoothness is important for domains such as model fitting and rule creation user interfaces, as explored in previous work. Goals for this project may include creating a unified growth language useful for multiple domains and creating user interfaces for growth-based modeling. Contact Tom.
Procedural modeling systems are typically constrained by a single computer’s memory. To alleviate this constraint, we may build procedural systems which operate on databases. Given different constraints on geometry evaluation speed, bandwidth, and storage, we may wish to evaluate procedural models at different times, at different scales. Contact Tom.
Given a parametric or procedural model we would like to be able to optimise it to improve physical properties ranging from structural stability to the ability to collect solar energy. However procedural models present complex search spaces, and physical simulation is slow. In this project we will examine deep neural architectures to accelerate this optimising process. Contact Tom.
In a past life Tom has a startup which edited video libraries together in the browser. The major outcome of this project was that i) no one wanted to pay for this, and ii) you could make very compelling videos by ransacing a video library for random half second clips and concatenating them together interactively. In this project we would evaluate this concept of micro-video from an user interface perspective, apply the systems to video comprehension tasks, and extend its use to different problems and domains. Contact Tom.