Honours and Masters project

In our current day and age, there is an exponential growth in multimodal data, especially the transition of social media from text-based communications to video formats which can be observed with the rise of TikTok, Youtube, and Instagram Reels. This shift requires a shift in how we analyze multimodal data as we will have to move away from traditional text sentiment analysis such as TextCNN. Multimodal data presents us the opportunity to improve on text-based analysis given the new information that is coded in speech and visuals that can provide additional context for a sentiment.

Podcasts have become a very popular way for small communities to create content that is meaningful for them and reach a wider audience. However, many of the skills and equipment needed to produce a good podcast are inaccessible to non-professionals and there is often a learning curve attached to gain necessary skills. In addition, the production process is seen as an individual effort (one or two producers working in isolation to produce the final edit).

Radio is one of the primary modes in which communities across the world receive important information and build connection with wider society. Non Governmental Organisations (NGOs) have long been leveraging radio, and in particular Community Radio

In many parts of the world, audio is the preferred interface for social interactions. There has been a huge push towards audio-based interfaces for engaging marginalised communities in rural and Developing countries contexts.

As part of this project, you will work closely with a community organisation or NGO (this can either be an organisation that you have existing links with or we will connect you with one of our partner NGOs). Working in collaboration with the org, you will find out challenges they face in giving voice to their communities/beneficiaries that can be addressed through social media (for instance, perhaps they want to run an awareness raising campaign about the difficulties faced by the community and they want the communities to be very involved in this).

Recently, program generation and optimisation techniques have been adapted to performance critical subroutines in cryptography. Codes generated/optimised by these techniques are both secure and their performance is highly competitive compared to hand-optimised code by experts [1].

IT Forensics is the art of extracting digital pieces of evidence also known as (aka) artifacts in a forensically sound manner, that is presentable to a court of law. In doing this it covers a range of conceptual levels, from high-level operating systems and computer theory down to computer networking.   

The specific objective(s) of this project is to look at an encrypted piece of data and distinguish what encryption algorithm is used/employed. This would benefit IT Forensics researchers/investigators attacking encrypted volumes, files, folders, etc.

Since the 1990s, researchers have known that commonly-used public-key cryptosystems (such as RSA and Diffie-Hellman systems) could be potentially broken using efficient algorithms running on a special type of computer based on the principles of quantum mechanics, known as a quantum computer. Due to significant recent advances in quantum computing technology, this threat may become a practical reality in the coming years. To mitigate against this threat, new `quantum-safe’ (a.k.a.

Mini-CP https://www.info.ucl.ac.be/~pschaus/minicp.html is a minimal form of constraint programming solver, designed to allow for easy experimentation and learning. 

One of the most efficient approaches to discrete optimisation solving is using lazy clause generation, which is a hybrid SAT/CP approach to solving problems.  But MiniCP does not currently support this. 

Given a knowledge base describing the existing background constraints and assumptions about what is possible in the world as well as the prior experience of an autonomous agent on the one hand and probabilistic perception of the current state of the world of the autonomous agent, on the other hand, it is essential to devise and efficiently enumerate the most consistent world models that are likely to be valid under the prior knowledge in order to refine the agent’s up-to-date perception and take the most suitable actions.

This project takes a different approach to RL, inspired by evidence that Hippocampus replays to the frontal cortex directly. It is likely used for model building, as opposed to the mainstream view in cognitive science and ML - where 'experience replay' ultimately improves policy. The predicted benefits are sample efficiency, better ability to generalize to new tasks and an ability to learn new tasks without forgetting old ones. The project objective is to improve biological models and advance state-of-the-art in continual reinforcement learning.