PhD Studentship: Tiny Multimodal Learning Under Resource Constraints Manchester Metropolitan University in United Kingdom

[Home tuition fees (£5,006 for 2025/26) covered for 3 years; international students must pay the difference. Standard UKRI stipend (£20,780 for 2025/26) for duration of award.]

Manchester Metropolitan University is offering a fully funded PhD studentship focused on Tiny Multimodal Learning Under Resource Constraints. This project addresses a critical challenge in modern artificial intelligence: how to achieve robust multimodal intelligence using minimal, efficient representations, especially in environments with limited computational, memory, or energy resources. Unlike traditional approaches that scale up model complexity, this research aims to develop principled, lightweight methods for representing and combining multimodal information such as audio, motion, images, and sensor signals.

The project is grounded in foundational machine learning, representation learning, and efficient algorithm design. Key objectives include benchmarking existing multimodal learning approaches under resource constraints, investigating lightweight data fusion strategies, studying the impact of temporal structure and missing data, and identifying trade-offs between accuracy, efficiency, and robustness. The research is highly relevant to sustainable and accessible AI, with potential applications in real-world systems where resources are limited.

Funding is available for both Home and International students. Home tuition fees (£5,006 for 2025/26) are covered for the full 3-year duration, and a standard UKRI stipend (£20,780 for 2025/26) is provided. International students are eligible but must pay the difference in tuition fees (Band 2 for 2025/26). The studentship is ideal for candidates with a strong academic background in Computer Science, Artificial Intelligence, Machine Learning, or closely related fields. Essential requirements include a first-class or upper second-class (2:1) Bachelor’s or Master’s degree, proficiency in programming (especially Python), and solid knowledge of machine learning or data analysis. Applicants should also demonstrate strong analytical, problem-solving, and communication skills, and be able to work both independently and collaboratively.

To apply, interested candidates should contact Liangxiu Han for an informal discussion. Formal applications require completion of the online application form for a full-time PhD in Computing & Digital Technology, submission of the Doctoral Project Applicant Form, and uploading of a CV and covering letter via the University’s Admissions Portal. The covering letter should clearly map your skills and experience to the project’s aims and objectives, and explain your interest in the research area. The application deadline is 16 March 2026, with an expected start date of 1 October 2026. Please quote the reference SciEng-LH-2026-27-Tiny Multimodal Learning in your application.