Collective Intelligence and Network AI

Federated learning is a paradigm breakthrough shift in AI for data privacy preservation. Unlike conventional AI techniques, federated learning does not require the users to share the data, but only the AI model parameters. As such, federated learning has found notable successes in numerous applications (e.g., Google keyboard, localization, security, and data sharing) and various disciplines … Read more

FPGAs to solve problems in current IoT devices

I am interested in supervising projects using FPGAs to solve problems in currentIoT devices: Power consumption and concurrencyDistributed and wireless acquisition of multimedia dataFPGAs to reduce the required data rate by preprocessing and data compression.Combinations of inexpensive sensors based on microcontrollers and FPGAs.Fast orthogonal transformations in signal/image processingCryptographic data protectionData compressionError-detecting codingFPGAs and Internet Protocol … Read more

Automatic tuning of logic synthesis optimization scripts [Taken]

ABC is an open-source logic synthesis and optimization tool developed by Alan Mishchenko at the University of California at Berkeley. ABC specializes in optimizing digital logic circuits at the circuit level, and some commercial hardware synthesis tools incorporate ABC as part of their design flow to improve their circuit optimization. ABC is a large logic … Read more

[Taken] OpenFlow Implementation for Information-Centric Networking

OpenFlow has been proposed as a protocol for managing devices in a control plane to direct the operation of switches in a data plane in order to adjust the flows of traffic within a network. This development of this protocol has focussed on IPv4/6 and other traditional associated protocols. Information-Centric Networking (ICN) – in contrast … Read more

Multi-objective Genetic Algorithm for Test Generation

Manual creation of tests to ensure the quality of software artefacts is a difficult and labour intensive task. Automatic test generation on the other hand (e.g., using a tool like Randoop) usually requires little or no help from the developers — but often ends up creating too many tests of little value. Various techniques have … Read more

A 3 Step Meta-heuristic for the Timetabling problem

Time tabling is a complex optimisation problem given its large number of constraints (e.g., no overlap between core modules for a particular year) or cost functions (e.g., minimise the distance between any two consecutive lectures/TA/labs).  Many techniques have been used to address various versions of the problem, such as, Constraint-based methods or Meta-heuristics methods, but … Read more

An AR/VR Rock Climbing Preparation App

Rock climbing is a demanding activity that often requires participants to study the route(s) they’ll have to follow before they start climbing. Thousands of books have been written to describe famous and less famous rock climbing routes, clubs (e.g., Club Alpin Français) propose seminars and workshops to prepare climbers etc. The objective of this project … Read more

[Unavailable] Cinematic Lighting in VR: Enhancing Emotional Storytelling Through Real-Time Illumination

[FYP or MSc Level] This project will explore how lighting design, informed by cinematography and psychological theories, can be used to enhance emotional expressivity in characters and user immersion within a real-time VR open-world environment. The student will research emotional lighting techniques from film and psychology—such as color theory, contrast, and lighting direction—and apply them … Read more