Skip to main content

Research projects in Information Technology

Displaying 1 - 10 of 179 projects.


Understanding material failure by machine learning analysis of pattern strains

Metals are made of small crystals - i.e., atoms are arranged in a particular geometric arrangement, which are typically in the range of a few 10s of microns (0.01 mm). The arrangement of these crystals greatly affects the performance of the metal and hence the performance of components where metals are used - such as in aeroplanes, gas turbine engines, cars, etc. The manner in which such materials deform, crack and fail under a variety of conditions is an important area in terms of cost and safety.

Spatiotemporal Object Detection/Segmentation from Multi-Spectral Time Sequence Data

Positions: We have an ARC fully-funded PhD project with generous top-up scholarship in the areas of machine learning and computer vision. The PhD project is 3.5 years, including at least a one-year equivalent industry placement, the timing of which can be negotiated.

Towards adversarially robust deep models

Deep Neural Networks have shown remarkable performance across a wide range of computer vision tasks. They are however vulnerable to carefully crafted, human imperceptible perturbations, which once added to the input images, can easily fool models' decisions. Such adversarial perturbations, therefore, pose a serious concern to the deployment of deep learning models in real-life scenarios. This project will aim towards developing reliable and trustworthy deep networks by e.g., exploring robust training strategies, loss formulations, and architectural modifications.  

Supervisor: Dr Munawar Hayat

Large Scale Zero Shot Object Detection

Zero-shot detection aims to simultaneously identify and localize (by predicting bounding box coordinates) objects which have never been observed during training time. The existing zero-shot detection approaches project visual features to the semantic domain for seen objects using textual embeddings learned in a stand-alone manner without any joint incorporation of image data. This project will aim to leverage from recent developments in joint image-text modeling, to find the more meaningful correspondence between visual features and their semantic embedding.

Supervisor: Dr Munawar Hayat

Machine learning for short message conversational analysis in Law Enforcement

This project aims to identify novel methods for inferring actors, activities, and other elements from short message communications. Covert communications are a specialist domain for analysis in the Law Enforcement (LE) context. In this project we aim to improve law enforcement’s understanding of online criminal communications, exploring texts for automated understanding of intent, sentiment, criminal capability, and involvement.

Supervisor: Dr Campbell Wilson

Explainability of AI techniques in law enforcement and the judiciary

This project will investigate and develop the ways in which AI algorithms and practices can be made transparent and explainable for use in law enforcement and judicial applications

Supervisor: Dr Campbell Wilson

Ethics of AI application in law enforcement

The use of AI in law enforcement and judicial domains requires consideration of a number of ethical issues.  This project will investigate and develop frameworks that embed ethical principles in the research, development, deployment ,and use of AI systems in law enforcement (LE). A major focus is expected to concern the acquisition, use, sharing and governance of data for AI in this context.

Supervisor: Dr Campbell Wilson

Using 3D Printing to Improve Access to Graphics by Blind and Low Vision People

This project seeks to explore the use of 3D printing to provide better access to graphical information to those who are blind or have low vision.

Supervisor: Dr Matthew Butler

Highlighting and Mitigating Various Biases in Big Language Models

In recent years, big language models such as GPT-2, GPT-3 and Switch-C with billions to trillions of parameters trained on 100s of gigabytes of text have enabled us to push the state-of-the-art performance in various tasks. But does training on web data come with no social or environmental cost? In this project we will highlight some of the inherent issues of these models in terms of various social biases they carry forward from data, and develop mitigating techniques to remedy these both at the level of data sanitisation and representation/modelling.

References

Workspace Layout Optimisation for Improved Operator Decision Making

Energy market operators make data-driven decisions via 24/7 control rooms with the use of many different applications across their multiple screen workstations. The types of decisions the operators are undertaking depend on the time of day and the state of the network. With the increase of data in recent years and the influx of distributed energy resources, the types of decisions and quantity of information needing to be looked at at a glance to make informed decisions is rapidly changing.

Supervisor: Dr Sarah Goodwin