27716 research outputs found
Sort by
Gaussian Process Methods for Target Tracking and Sensor Scheduling
No full textThis thesis explores the application of Gaussian Processes to practical target tracking
problems. It begins with the development of a general-purpose codebase for implementing
Gaussian Processes, creating a flexible foundation for various tracking scenarios. To tackle
the computational challenges inherent to Gaussian Processes, the research investigates
several sparse Gaussian Process methods. These methods reduce computation time while
preserving accuracy, demonstrating significant improvements in both inference speed and
tracking performance.
The study advances these foundational methods by integrating variance-based Bayesian
optimisation techniques to enhance tracking and sensor management for Unmanned Aerial
Vehicles. This approach utilises the probabilistic nature of Gaussian Processes to optimise-
mize resource deployment dynamically, boosting target localisation and tracking efficiency.
Beyond continuous tracking, the research extends to Gaussian Process Classification to
address discrete decision-making scenarios. These additions highlight the versatility of
Gaussian Processes in Target Tracking Applications.
The results confirm that Gaussian Processes, when combined with sparse approximations-
tions and Bayesian optimisation, offer an effective and scalable framework for improving
target tracking performance. By focusing on dynamic and adaptive methods, this research
contributes to the development of tracking algorithms that meet the demands of diverse
and evolving scenarios
Developments of High-speed Near-infrared Fourier domain Mode-locking Optical Coherence Tomography for Clinical Applications
Full text linkOptical coherence tomography (OCT) has gained significant attention in both medical and non-medical fields due to its non-invasive, real-time, three-dimensional, in-vivo, and multi-functional imaging capabilities. In dermatology, OCT enables the visualisation of tissue microstructure, vasculature, and collagen distribution, making it highly effective for diagnosing and monitoring conditions such as skin cancers and inflammatory diseases. However, the slow scan rates of commercial systems limit the acquisition speed of functional information.
This thesis addresses that challenge through the development of a high-speed near-infrared Fourier-domain mode-locked (FDML) OCT system tailored for clinical dermatological applications. Chapter 2 presents the system, which achieves an ultrahigh A-scan rate of 1.67 MHz, with axial and lateral resolutions of ~13 µm and ~25 µm respectively, sensitivity of ~100 dB, and imaging depth of ~4.8 mm in air. Real-time imaging and data acquisition are enabled through custom software, and the system’s capability is demonstrated through in-vivo imaging of various skin sites.
Chapter 3 demonstrates rapid measurement of skin biomarkers, including epidermal thickness and vascular morphology, using custom automated segmentation and OCT angiography algorithms. Chapter 4 presents the first application of variable interscan time analysis (VISTA) for quantifying cutaneous blood flow using an FDML OCT system. A novel auto-fitting method is also introduced into the VISTA framework, improving the accuracy of decorrelation coefficient estimation.
Chapter 5 proposes a novel algorithm for skin topographical matching to enable precise re-scanning of treated sites. This is the first OCT-based approach that automatically estimates matching scores to support longitudinal skin monitoring. Additionally, this FDML OCT system is combined with an in-house designed colposcopic scanning probe to aid in its alignment and characterisation.
Finally, Chapter 6 summarises the system's development and discusses its potential applications and future extensions. This high-speed FDML OCT system demonstrates readiness for translation into dermatological clinics
Designer-artisan co-design for value co-creation: Case studies of the Chinese traditional textile sector
No full textTraditional crafts are vital cultural and economic resources that support local societal sustainability. However, rapid industrialisation, globalisation and misaligned design interventions have led to their decline. While existing studies underscore the importance of designer–artisan co-design, it remains unclear how to optimise and sustain long-term co-design capabilities in the presence of complex social dynamics.
To address these challenges, this research proposes a systematic model for developing designer–artisan co-design in the China’s traditional textile sector, focusing on bridging the macro, meso and micro levels of co-design practices.
This study developed a co-design process model that incorporates interconnected components across stakeholder levels. It demonstrates how social contexts interact with the partners’ co-design competitiveness and capabilities. The model also illustrates a co-design process between designers and artisans, incorporating interconnected co-design strategies to balance diverse capabilities while enhancing the craft product design quality.
Moreover, the practicality of key aspects of existing designer–artisan co-design tools and toolkits in real-world settings was evaluated. This research’s findings may guide interconnected approaches to long-term designer-artisan co-design across different textile sectors and social contexts. The findings also provide insight into holistic co-design theory for societal sustainability and practical multilevel co-creation practices, which stakeholders can use to develop transformative, interdependent tools and support systems. By linking academic theory with practical application, this study promotes sustainable innovation in traditional craft industries for overall societal development and offers valuable insights for researchers and practitioners seeking to revitalise traditional textile sectors
Exploring socio-technological-ecological dimensions of Asian songbird trade.
No full textSince its inception, conservation has grappled with the ‘problem’ of wildlife trade, often overlooking the experience of animals, plants, and humans involved in the trade chain. By integrating political ecology, machine learning, and design theory, this thesis redefines wildlife trade as a multifaceted issue, or a ‘wicked problem’. The thesis focuses on the trade of live songbirds as pets in Indonesia, a trade deeply embedded in culture, monetized, diverse, and often illegal. Given increasing trade, the International Union for the Conservation of Nature termed the phenomenon of over-harvesting birds as the Asian Songbird Crisis (ASC). I examine the premise that species identification is a significant barrier to combating the illegal songbird trade in Indonesia and explore the development of taxonomic aids for conservation law enforcement. I show that image-based machine learning shows promise for identifying birds in complex trade environments (Chapter 2). However, focusing on a single solution, like a taxonomic aid, may solve some issues (species identification) while exacerbating others (enforcement challenges). Enforcement mostly focuses on community-based policing and education for vendors (socialisasi) (Chapter 3). In Chapter 4, I explore user-centered design of taxonomic aids with conservation law enforcement (CLE). Barriers to technology adoption include the social acceptability of bird trade, corruption, and the need for verbal evidence in court. There are diverse attitudes towards the ASC, with many CLE actors not recognizing the crisis label. Market mechanisms shift harvesting pressure onto similar species, suggesting a sustainability-focused narrative may be more productive (Chapter 5). Chapter 6 develops a more-than-human political ecology approach, exploring the interconnectedness between people and birds in trade. This chapter highlights previously obscured harms to birds, such as feather plucking, dismemberment, and overcrowding. Future research should explore alternatives to punitive measures to address unsustainable wildlife trade embedded within complex cultural practices
The Perspectives of Teaching Staff on Institutional Change Management in the Implementation of Online Learning
No full textThis study took place at a Higher Education institution in the UAE during a time of pedagogical change from face-to-face teaching to fully online instruction. Over a period of two years, the practices and cognitions of four teachers were studied through a qualitative case study methodology using interviews and classroom observations. Theoretically, the study draws on teacher cognition research as well as two influential models relevant to technology use in education: TPACK (Technological Pedagogical Content Knowledge) and UTAUT (Unified Theory of Acceptance and Use of Technology). These frameworks provide a conceptual scaffold for comprehending the complex and dynamic relationships involved technology-driven pedagogical change. The findings of this study indicate that teachers’ cognitions, institutional processes, and student factors play key roles in the effective implementation of pedagogical change. Throughout the case study, emphasis is placed on the importance of considering educators' cognitive processes during the implementation of pedagogical innovations. Teachers’ attitudes, beliefs, experiences, and perspectives inform their instructional choices and acceptance of policy. Challenges identified by the participants during the change process included institutional preparedness, student responsiveness, and the need for further professional development. Effective teacher facilitation was found to be key to the successful implementation of online teaching, in addition to the need for specifically designed learning materials. Institutional change management is also an important influence on the successful implementation of innovative pedagogical approaches. This study acknowledges that educational change requires more than a policy directive and the incorporation of novel technological tools; it entails a consideration of the nuanced interplay of cognition, context, and communication. These findings contribute to the broader dialogue on teacher cognition during periods of educational change, especially that which is technology-mediated, and highlight the significance of teacher agency, contextual appropriateness, and institutional preparedness
Neural Network-Based Surrogate Models of Lagrangian Continuum Simulators
Full text linkThe simulation of fluid and deformable solid dynamics is a cornerstone in disciplines ranging from engineering and environmental modeling to computer graphics and virtual reality. Traditional computational methods, despite their power, often struggle to balance accuracy, efficiency, and scalability, particularly in real-time and large-scale applications. This thesis addresses these challenges by developing neural network-based surrogate models specifically tailored for Lagrangian continuum simulations.
The research explores temporal learning within continuum simulations, starting with an in-depth analysis of encoding and decoding techniques that transform Euclidean coordinate-based continuum data into latent space representations. This transformation is essential for training neural networks to accurately model complex physical phenomena. The study systematically compares traditional time series prediction methods with advanced neural architectures, such as Long Short-Term Memory (LSTM) networks. Experimental results demonstrate that LSTM networks, when combined with Multi-Layer Perceptrons (MLP), significantly outperform traditional methods in capturing the intricate multi-material interactions and long-term dependencies inherent in Lagrangian simulations.
A key contribution of this research is the introduction of a Self-Supervised Graph Attention Operator, which enhances the neural network’s ability to capture and conserve critical physical properties like vorticity and energy across simulated particles. This operator enables accurate and stable predictions of complex fluid and deformable solid dynamics, overcoming the limitations of existing surrogate models that often compromise between computational efficiency and physical accuracy. Rigorous evaluations, including long-term stability tests and comparative analyses with traditional methods, demonstrate that the proposed models advance the state-of-the-art in surrogate modeling for continuum simulations
Electrical and Optical Properties of Femtosecond Pulsed Laser Deposited Vanadium Dioxide Thin Films
No full text“Never a right time, never a right place”: Chinese single working women’s work-life experiences and negotiation of career identity
No full textThis research investigates how the status of singlehood and childfree affects Chinese single working women’s work-life experiences and their sense of self related to their career, under the current pro-marriage and pro-natal social culture. The government’s keen desire to boost population sustainability, reflected by emphasising women’s ‘central role in the family’ in 2013 and the abolishment of the One-Child Policy in 2016, renders women’s effort in the workplace less important than their ‘obligation’ to become wives and mothers. Such advocacy of traditional gender division of labour leads to women facing intensifying discrimination and marginalisation in their daily work and the process of job seeking.
Previous studies in gender and employment suggest that women’s fulfilment of gender expectations through role transitions to spouse and motherhood usually leads to further gender-based disadvantage and inequality, such as the motherhood penalty. Nevertheless, whether the social pressure from women’s unfulfilled gender expectations affects their employment and work experiences is a question worth further investigation. In contemporary Chinese society, where single women beyond their mid-20s are usually labelled as ‘leftover women’, with an accusatory meaning that they miss the right time to marry because of educational or career pursuits, the question of gender expectations remains underexplored. Through 64 biographical narrative interviews with Chinese single working women located in mainland China, the United Kingdom, and the United States, this research argues that the career experiences and self-perception at work of single women in China are affected by the intersectional pressure from gender, age, and singlehood. For Chinese single women in the UK and US, their immigrant status interplays with their gender and singlehood, which consists of a new set of intersectional pressures which also affect their work and employment. The theoretical contribution of this research focuses on underestimated and undertheorized difficulties faced by Chinese single women between productive and reproductive roles at work and life interface imposed by traditional gender role expectations derived from Confucianism and neoliberal value orientation brought by marketisation. In this thesis, I principally argue that single working women of reproductive age in China are disadvantaged by the intersectional pressure of their gender, age, and singlehood in the labour market and workplace. Compared with the work and life stories of single women living and working in the UK and the US, I further argue that these employment disadvantages related to women’s single and childfree status at reproductive age are culturally bound to China’s contemporary social context. I propose a notion of “double penalty” to address the unique dilemma single working women face in China’s labour market and workplace, in which their single and childfree status is simultaneously translated into the absence of wifehood and motherhood at present and transition into wifehood and motherhood in soon future, disadvantaging them synchronously
Understanding Coercive Control Within and Beyond the Domestic Environment
Full text linkIn England and Wales, 2015, the Serious Crime Act (section 76) implemented the criminalisation of controlling or coercive behaviour, closing a gap in the law around ‘patterns of abuse’ in intimate partner or family relationships. Coercive control is at the heart of most intimate partner/domestic abuse, however, conviction rates have remained low. This study therefore investigates why evidencing/seeking support for coercive-controlling offences remains a challenge. Founded upon feminist standpoint epistemology, the study was conducted via 20 semi-structured interviews: ten with survivors of coercive control; one with the parent of a 12-year-old survivor; nine with practitioners/professionals working with victim-survivors and perpetrators of coercive control. Experiences explored in these interviews are analysed using thematic methods, providing rich insights into coercive control and building on how theory/practice can take the concept of coercive control forward.
My research shows the definition of coercive control informing UK policy/service intervention is limited in its confinement to the domestic. Dominant conceptualisations fail to capture victim-survivors’ experiences of coercive control that extend, structurally and/or systemically, beyond the domestic, across services intended to help, and enabling perpetrators to continue offending. This problem is also identified in relation to systematic patterns of abuse, perpetrated within workplaces, communities, and families, demonstrating how coercive control occurs both interpersonally and collectively: within and beyond the domestic on a continuum extending into organisational/institutional help-seeking contexts. My research addresses three key areas of coercive control in policy and practice: difficulties evidencing coercive control by criminal justice and support systems; how private/interpersonal and public/social/institutional violence intersect in coercive control; coercive control operating on a continuum. I argue existing conceptualisations and policy practices struggle to recognise and respond to the complex spectrum of coercive control; therefore proposing a framing not limited to domestic/interpersonal but operating on a continuum between private/public, individual/collective, domestic/institutional, meaningfully extending beyond the domestic
The role of subcutaneous adipose tissue in the pathogenesis of diabetes and cardiovascular disease
No full textAdipose tissue dysfunction is central to the heightened cardiovascular risk observed in people with type 2 diabetes mellitus (T2DM), yet the biological mechanisms underpinning this remain poorly defined. This thesis presents a series of epidemiological and translational studies suggesting that the microvascular biology of subcutaneous adipose tissue (SAT), particularly endothelial senescence, is critical in driving adverse cardiometabolic phenotype and outcomes.
Using data from half a million people from the UK Biobank, I demonstrate that obesity and diabetes are independently and additively associated with more advanced cardiovascular disease and increased incidence of major adverse cardiovascular events. These outcomes only weakly correlate with conventional risk markers, underscoring the need for deeper mechanistic insights. To explore this, I perform detailed phenotyping of SAT and primary human SAT microvascular endothelial cells (hSATMVECs) from individuals with heart failure and/or T2DM. These studies show that SAT in this population is fibrotic with impaired angiogenesis and characterised by pronounced endothelial senescence. I go on to show that senescent hSATMVECs exhibit dysfunctional crosstalk with adipocytes, likely mediated by the senescence associated secretory phenotype which induces a pro-inflammatory adipocyte phenotype resistant to glucose uptake.
Finally, I demonstrate that the cardiac glycoside digoxin, at a clinically therapeutic concentration, reduces hSATMVEC senescence, improves healthy hSATMVEC–adipocyte crosstalk, and improves adipocyte function. These findings provide a strong foundation for the further investigation of SAT microvascular senescence as a therapeutic target in T2DM and suggest potential utility for repurposing digoxin as a novel senolytic therapy in this context
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇