Dr Wenbin Li
Wenbin is interested in developing unified autonomous systems across various robots and their applications in manufacturing, professional filming, and search-and-rescue. Such research involves a range of topics in machine learning theory and engineering efforts, including reinforcement learning and its inverse problems, graph and network optimisation, and continuous state approximation.
Research interests:
University of Bath Research Portal:
https://researchportal.bath.ac.uk/en/persons/wenbin-li
Dr Georgios Exarchakis
Georgios finds great satisfaction in seeing well-defined theories validated through real-world observations. With a background in mathematics, he became fascinated with developing models to understand brain function. Over the past decade, machine learning has grown rapidly, driven by breakthroughs in deep learning that solved problems once thought unsolvable. This raises the critical question: Why were our expectations so misaligned? In machine learning, complex models often mimic human processes, but their function cannot be fully specified. Georgios’ research focuses on uncovering the underlying conditions that make these algorithms work—particularly why neural networks generalize so effectively across diverse domains. His work centers on representation learning and probabilistic modelling, and it has found application in diverse areas such as astronomy, medical imaging, and quantum chemistry, with an emphasis on understanding the mathematical principles behind these models’ success.
Research Topics:
Georgios Exarchakis — the University of Bath’s research portal
Dr Rohit Babbar
Rohit is a Senior Lecturer in the department of Computer Science at the University of Bath. Before this, he was an Assistant Professor at Aalto University in Finland, and a post-doc at Max-Planck Institute for Intelligent Systems, Germany.
The focus of his current research is on : (i) supervised learning problems with large label spaces, (ii) learning with long-tailed data, and under label/input noise, (iii) sparse neural networks for energy & memory efficient training/fine-tuning with LLMs, and (iv) generalisation/robustness in deep learning. Some of his works on developing memory efficient algorithms on commodity GPUs have been published in ECML 2023 and NeurIPS 2024. Other works on proposing evaluation metrics for long-tail data in large output space, their statistical properties and optimal classifiers have been published in ICLR 2024, NeurIPS 2023, and KDD 2022.
Research interests :
Rohit Babbar — the University of Bath’s research portal
Dr Michael Yang
Michael’s research is in the fields of Visual Computing and Computer Vision with specialisation on Scene Understanding, Multimodal Learning, Deep Generative Models. Scene understanding involves enabling computers to interpret and analyse visual data, such as images or videos, to recognise objects, people, and activities within complex environments. One recent research focus is on developing AI models for 3D scene synthesis, a rapidly growing field driven by the demand for realistic virtual environments in applications such as gaming, AR/VR, and robotics. The goal is to push the boundaries of current techniques to improve the realism, diversity, and practical usability of generated 3D scenes for real-world applications.
Michael Yang — the University of Bath’s research portal
Dr Ali Uncu
I work on developing and applying formal and human-verifiable symbolic computation algorithms to prove novel mathematical results. I am particularly interested in enumerative and algebraic combinatorics and number theory, especially from the problems arising from q-analogs and the theory of partitions. I am just as interested in the utilization of SAT/SMT (satisfiability and satisfiability modulo theories) methods and applied algebraic geometry to solve/simplify mathematical problems regardless of the source.
Research interests
Ali Uncu — the University of Bath’s research portal
Dr Alessandro Leronni
My research is centred on the continuum modelling and finite element simulation of complex multiphysics systems, with the aim of elucidating the mechanisms that dictate their behaviour and failure. My primary expertise lies in electrochemo-mechanical systems, and my work spans across various domains, including (1) smart materials, (2) energy storage, and (3) developmental biology. Specifically, I am developing multiphysics frameworks to investigate (1) the cold sintering of electroactive ceramic composites for actuation, sensing, and energy harvesting, (2) the micromechanics of lithium dendrites that propagate in solid-state batteries eventually causing their failure, and (3) the interaction between bioelectrical and mechanobiological signalling in cancer progression and tissue regeneration.
My research is inherently multidisciplinary, integrating principles from solid mechanics, fluid mechanics, and electrochemistry, and it often involves collaboration among engineers, mathematicians, physicists, and biologists.
University of Bath Research Portal
https://researchportal.bath.ac.uk/en/persons/alessandro-leronni
Dr Da Chen
I am currently a Lecturer in the Department of Computer Science at the University of Bath. Prior to this, I was a Postdoc researcher in a joint project between Alibaba Group and The Institute of Automation, Chinese Academy of Sciences.
My research focuses on computer vision, machine learning, multimodal learning, and related areas. I am particularly interested in solving complex practical tasks under “limited” data conditions, such as learning with limited labelled data, out-of-distribution data, cross-domain learning, incremental learning, etc. To this end, I have proposed multiple solutions for few-shot learning image classification, out-of-distribution detection, and other tasks. I am also interested in video-related computer vision and multimodal learning tasks such as video summarization, video object detection, dense video captioning, etc.
https://researchportal.bath.ac.uk/en/persons/chen-chen-2
Dr Antonio Pellegrino
My expertise lies in developing innovative experimental techniques to measure the mechanical response of engineering materials at high strain rates. I focus on understanding how variables such as strain rate, temperature, and environmental conditions affect lightweight reinforced polymers, titanium alloys, and cellular materials. Recently, I have developed a piezo-driven methodology for the synchronisation and timing of concurrent stress waves. The system allows for the synchronisation of tensile and shear stress waves and for the synchronisation of stress pulses in multiple directions.
I am also interested in applications of artificial intelligence in solid mechanics. I am actively involved in developing data-driven constitutive models to predict the mechanical behaviour of materials subjected to complex thermo-mechanical loading paths. These models account for various factors such as temperature, loading history, strain rate, and stress state, contributing to a deeper understanding of material behaviour in real-world scenarios.
Research Interests:
https://researchportal.bath.ac.uk/en/persons/antonio-pellegrino
Professor Jun Zang
Jun’s research concerns the hydrodynamic loadings on urban, coastal and offshore structures and the impact of extreme events on such structures. Her research group develops and uses advanced CFD tools in modelling urban flooding, coastal flooding, fluid structure interactions, performance and survivability analysis of marine renewable energy devices (including all wave, tidal and offshore wind energy) and violent wave impact on coastal and offshore structures.
LINKS:
Jun Zang on University of Bath Research Portal
