Respiratory diseases such as COVID-19 are known to physically damage our airway and lungs, which in turn alters the produced respiratory sounds (e.g., cough, breadth). During the pandemic, cough classification has emerged as an accessible, low-cost, and environmentally friendly COVID-19 screening alternative, needing only a smartphone to collect and process cough samples. However, audio processing … full description “AI-powered COVID-19 detection via cough sounds (available)”
Project List
Distilling large language models (available)
Model distillation for large language models (LLMs) presents a key challenge in AI research: how to compress massive, computationally expensive models into smaller, more efficient versions while preserving their performance. Large models, such as GPT-4, require vast amounts of memory and processing power, making them impractical for real-time applications on edge devices or personal computers. … full description “Distilling large language models (available)”
Exploring optimization algorithms for deep sequence models (available)
Deep sequence models such as recurrent neural networks (RNNs) are a key type of architecture in modern deep learning, particularly for processing sequential data such as language text, speech, video, and time series data. RNNs have loops that allow information to persist and be passed from one step to the next, enabling them to effectively … full description “Exploring optimization algorithms for deep sequence models (available)”
Implementation of biologically inspired efficient deep learning models (available)
As deep learning models continue to grow in size and complexity to tackle increasingly difficult tasks, the need for efficient and scalable models becomes ever more important. Extremely large language models like GPT-4 require massive computational resources and expensive hardware to train and run. This makes them impractical to deploy at scale in many real-world … full description “Implementation of biologically inspired efficient deep learning models (available)”
A Comparative Analysis of Guardrail Frameworks for Large Language Models (available)
This project explores how specialised programming frameworks called Guardrails, developed specifically for constraining large language models (LLMs), can prevent them from generating harmful, biased, or off-topic content. The goal of the project is to build simple examples using three leading guardrail frameworks implemented in Python: Guardrails AI, NeMo Guardrails from NVIDIA, and Llama Guard from … full description “A Comparative Analysis of Guardrail Frameworks for Large Language Models (available)”
A Cyclic Prover for Propositional Dynamic Logic (available)
Propositional Dynamic Logic (PDL) [1] is a logic in which abstract properties about the behaviour of programs can be expressed, in a very general way. Because of its generality it is suitable for checking the behaviour or a wide range of programs and systems, via model checking. PDL is what is known as a modal … full description “A Cyclic Prover for Propositional Dynamic Logic (available)”
A Debugging User Study: Validating Debugging Research Findings (available)
Project Description: Understanding and fixing software faults is a challenging task for developers [1]. To address this challenge, researchers have designed several debuggers. For instance, automated fault localisation (AFL) techniques (e.g., Ochiai [1]) and automated program repair (APR) tools (e.g., GenProg[2]) are designed to support developers during software debugging tasks. In addition, researchers have gathered … full description “A Debugging User Study: Validating Debugging Research Findings (available)”
Algorithms for lifelong deep learning (available)
Continual learning [1], also known as lifelong learning, refers to the ability of an artificial intelligence system to continuously learn and adapt from new experiences over time. This is an important capability as it allows AI models to acquire new knowledge and skills as more data becomes available, without forgetting previously learned information. Continual learning … full description “Algorithms for lifelong deep learning (available)”
Analyzing the Evolution of Fairness properties in ML-based Code repositories (available)
Project Description: Machine Learning (ML) systems are vital components of everyday products and services in our society. ML models have been popularly employed in law enforcement (e.g., COMPASS), legal use cases (e.g., LegalBERT), and programming language tasks (e.g., CodeBERT, Github Copilot, etc.). Despite the criticality of these use cases, ML systems are often biased towards … full description “Analyzing the Evolution of Fairness properties in ML-based Code repositories (available)”
Are we alone? Discovering Earth-like exoplanets with Conformal Prediction (available)
Billions of exoplanets are orbiting around their stars outside our solar system [1]. But are they similar enough to our Earth so that life may have developed there? Often you can answer this question if you know the planet’s mass, radius, and orbiting period. Given the astronomic distances between us and the planets, however, you … full description “Are we alone? Discovering Earth-like exoplanets with Conformal Prediction (available)”
Automated Debugging of Invalid Inputs generated by Fuzzers (available)
Project Description: Fuzzing is a popular testing method used to ensure the reliability, security and correctness of software systems. These tools allow developers to find bugs and vulnerabilities in software systems automatically. For instance, AFL is a popular fuzzer that has exposed thousands of bugs in open-source software provided by Google, Amazon and Firefox [1]. … full description “Automated Debugging of Invalid Inputs generated by Fuzzers (available)”
Autonomous Cars – Evaluation of Security Countermeasures (available)
Project Description Autonomous and connected vehicles will be a part of reality in the near future. There are many development efforts currently underway to pave the way for the deployment of autonomous vehicles (self-driving cars) in public areas. These vehicles are a collection of complex and sophisticated computational architectures. Cybersecurity is among many challenges that … full description “Autonomous Cars – Evaluation of Security Countermeasures (available)”
Carbon Labelling – Blockchain based product carbon footprint system. (available)
Project Description Carbon footprint associated with a product is a combination of the emission-related in the manufacturing and logistics of the product. Logistics plays a vital role in moving products from factories to home and is one of the significant causes of carbon emission in Europe. Besides this, the last mile delivery network is also … full description “Carbon Labelling – Blockchain based product carbon footprint system. (available)”
Computer Vision for Extreme Environments (available)
The use of data from extreme environments in computer vision have shown an increase of interest in recent years as drones and autonomous vehicles were introduced into new uses. Nuclear plants, deep underwater and space vehicles are some of the areas computer vision can be applied to develop a fully autonomous system. Furthermore, the development … full description “Computer Vision for Extreme Environments (available)”
Corporate Transparency – Enabling Consumers Track and Vet their Data Usage (available)
Project Description Data, primarily related to consumers is a crucial part of the modern tech-companies like Google, Facebook, Instagram, and non-tech-companies (superstores, insurance companies, etc.) alike have now started to rely heavily on consumer data for various commercial activities. Although the General Data Protection Regulation (GDPR) enhances the rights of the consumers from a technical … full description “Corporate Transparency – Enabling Consumers Track and Vet their Data Usage (available)”
COVID-19 – Mobile phone based passport applications (available)
Project Description We have all been affected by COVID19. Amongst the main driving forces for reopening international borders and enabling traveling is the existence of COVID19 immunity passports. However, there are serious considerations around the overall trust, security, privacy but also ethical dimensions of such proposals. This project aims to take into account the views … full description “COVID-19 – Mobile phone based passport applications (available)”
Cybercrime and ransomware groups – data analysis (available)
Project background Cybersecurity is the practice of defending computers, servers, mobile devices, electronic systems, networks, and data from malicious attacks, in an ever increasingly complex threat landscape. These attacks constitute a variety of computer-enabled and computer-dependent crimes, broadly categorised as ‘cybercrime’. In order to be in a position to defend against these attacks and minimise … full description “Cybercrime and ransomware groups – data analysis (available)”
Cyclist: Verifying Programs using Cyclic Proof (available)
Cyclic proofs are a technique for proving properties inductively, or proving properties about inductively defined objects or processes. They can be used to prove logical statements, or to verify recursive programs. The Cyclist tool [1] has been developed as an automatic verification tool based on cyclic proofs. It can prove termination and verify temporal properties … full description “Cyclist: Verifying Programs using Cyclic Proof (available)”
Deep learning based underwater image segmentation and object recognition (available)
In the past decade, advances in marine object recognition have been dramatically boosted for monitoring of underwater ecosystems. Traditional statistical analysis and ocean model simulation heavily depend on the availability of visual features. However, Due to light attenuation and scattering problem, the underwater images captured by optical imaging system are heavily degraded. As a result, … full description “Deep learning based underwater image segmentation and object recognition (available)”
Deep Learning-based Environmental Sound Classification (available)
Automatic sound classification attracts increasing research attention owing to its vast applications, such as robot navigation, environmental sensing, musical instrument classification, medical diagnosis, and surveillance. Sound classification tasks involve the extraction of acoustic characteristics from the audio signals and the subsequent identification of different sound classes. In this project, we will explore diverse deep neural … full description “Deep Learning-based Environmental Sound Classification (available)”
Detecting money laundering in cryptocurrency economy (available)
Project Description Cryptocurrency networks are unique in a way that all the transactions are public. All payers and receivers are known, and their respective pseudonyms are public. Whereas the inherent anonymity enabled by the pseudonyms and privacy-preserving protocols, cryptocurrency can be argued to be a suitable platform for money laundering activities. In this project, the … full description “Detecting money laundering in cryptocurrency economy (available)”
Embeded programming for INTI (available)
We are building a low-power transputer (parallel computer with microcontrollers) to study neuromorphic computing and run inference in low power devices (low power as in low computing power and low energy consumption). It is named the Incipient Neuromorphic Transputer Initiatiave in honour of Inti the Inca sun god. And because it is an incipient effort … full description “Embeded programming for INTI (available)”
Empirical evaluation of static verifiers for the Go programming language (available)
Go is a language that natively supports many constructs to synchronise concurrent threads. Because of this, developing code that contains subtle concurrency bugs is rather common. To address this problem, several groups of researchers have put forwards tools to find concurrency bugs in Go codebases. The objective of this project is to evaluate how some … full description “Empirical evaluation of static verifiers for the Go programming language (available)”
Exploring Partitioning for Conformal Binning Strategies (available)
Conformal Binning is a technique, inspired to traditional Conformal Prediction, and very recently proposed in the literature (https://arxiv.org/html/2503.03841v1) which involves partitioning the data into bins based on certain characteristics, such as the range of predicted values or difficulty estimates. Separate prediction intervals are then constructed for each bin, tailored to the specific distribution of data … full description “Exploring Partitioning for Conformal Binning Strategies (available)”
Federated Machine Learning – Security and Privacy Evaluation Framework (available)
Project Description User data is essential for many of the modern business operations, especially related to the building consumer segmentation and profiling – for marketing and services personalisation. Dealing with user data has its positive and negatives, especially after the General Data Protection Regulation (GDPR) collecting and storing user’s personal data. Centralised machine learning approaches … full description “Federated Machine Learning – Security and Privacy Evaluation Framework (available)”
Formalising Recursive Adaptive Grammars using a Proof Assistant (available)
The aim of this project is to formalise using proof assistant (such as Coq [1], Agda [4], or Isabelle [5]) the meta-theory of the recursive adaptive grammar framework [2] Programming languages are usually specified on two levels. The first level defines a context-free grammar that specifies the syntax of the language. Usually, however, not all … full description “Formalising Recursive Adaptive Grammars using a Proof Assistant (available)”
Intelligent Forensics for Trusted Execution Environments in RISC-V Architectures (available)
Project Description: Trusted Execution Environments (TEEs) offer a promising approach to securing computation on embedded systems, but digital forensics capabilities within these environments remain underexplored. This project investigates how Machine Learning (ML) and Artificial Intelligence (AI) techniques can be integrated into a lightweight digital forensics framework tailored for TEEs, with a focus on RISC-V-based systems … full description “Intelligent Forensics for Trusted Execution Environments in RISC-V Architectures (available)”
Jupyter notebooks (available)
Prerequisites: experience with Python (useful) and Javascript programming (essential). It may be useful to be willing to learn about functional programming (but this is not essential). Jupyter notebooks [1] are examples of literate programming [2] where code and outputs from the code as well as documentation are in the same application. Jupyter allows the user … full description “Jupyter notebooks (available)”
Linux Kernel Driver Development in Rust (available)
Rust is a new programming language for performance-critical systems software, e.g. operating system kernels, hypervisors, databases, and web servers [1][2]. As with other systems programming languages (e.g. C/C++), Rust provides low-level control over hardware resources (e.g. memory layout and allocation). In contrast to C/C++, Rust’s strict compiler checks help prevent common programming errors, reducing the … full description “Linux Kernel Driver Development in Rust (available)”
Maritime Cybersecurity (available)
Project Description Connected objects in motion are developing as a significant domain. There are many traditional businesses, like logistics and maritime shipping that have legacy systems, but they are rolling out interconnectivity globally. Today’s shift towards increasing interconnectedness at sea is continuing to enable significant efficiency gains and new capabilities for maritime operations. Running in … full description “Maritime Cybersecurity (available)”