University of Birmingham
University of Birmingham Research Archive, E-theses RepositoryNot a member yet
11811 research outputs found
Sort by
Towards more flexible and efficient RGBD object tracking
Get PDFObject tracking is a fundamental task in the area of computer vision. Recently, RGBD (RGB+Depth) object tracking has gained lots of attention due to the development of depth cameras. Compared to RGB-only tracking, RGBD tracking opens more opportunities for accurate and robust object tracking in complex scenarios, such as background clutter and dark scenes, as depth clues are helpful on object and background interference, especially in color failed occasions. However, the development of RGBD tracking severely lags behind its RGB counterparts and remains far from sufficient for real-world applications due to various challenges. On the one hand, current RGBD tracking is limited by 2D settings, which constrains RGBD tracking on 2D bounding box descriptions and neglects its potential for flexibility brought by depth information. On the other hand, the efficiency of RGBD trackers is ignored, which impedes the realistic applications of RGBD tracking. This thesis addresses the aforementioned issues and contributes to more flexible and efficient RGBD object tracking. In particular, a series of works are presented to explore and demonstrate the power of RGBD tracking.
The four main contributions of this thesis are:
The first contribution of the thesis introduces a novel paradigm, i.e., weakly-supervised RGBD video object segmentation, which achieves pixel-level RGBD object tracking under weak supervision. By exploring robust cross-modal fusion, the proposed FusedCDNet performs RGBD tracking on pixel level with only bounding box level supervision in both training and testing.
The second contribution of this work introduces generic 3D object tracking in RGBD videos. Specifically, a novel Track-it-in-3D dataset is proposed with rotated 3D BBox annotation, which bridges the gap between RGBD tracking and point cloud tracking. Also, a strong baseline is given for generic 3D object tracking with color and depth fusion and 3D-level cross-correlation.
The third contribution presents a study on the training-efficient tracking paradigm, which addresses the high training cost problem in RGBD tracking by applying a prompt learning paradigm. With the proposed cross-modal prompts, both the large-scale RGB knowledge from pre-trained large models and complementary information from depth sensors can be well explored. Moreover, the prompting framework is effective on different multi-modal object tracking tasks and its effectiveness is verified on different multi-modal tracking scenarios, including RGB+D, RGB+T, and RGB+Event tasks.
Finally, we present an efficient and lightweight approach for RGBD tracking, which is the first study on efficient RGBD object tracking. With efficient modality-aware fusion and lightweight backbone, the proposed RGBD tracker EMT runs at a speed of over 100fps. We also provide on-board scenarios and newly defined overhead space for RGBD aerial tracking. In such a scenario, many more categories (34 classes) can be considered for multi-modal aerial tracking than existing aerial tracking datasets. The corresponding on-board tests demonstrate that the proposed EMT can achieve real-time tracking on edge platforms.
As outlined, it is concluded that, by fully exploiting depth clues for solving the problems in current RGBD tracking, more flexible and efficient RGBD object tracking can be achieved
Manufacturing and development of implantable drug delivery device for the localised treatment of glioblastoma
Get PDFGlioblastoma (GBM) is an aggressive brain tumour with low survival rates. The blood-brain barrier (BBB) limits drug penetration into brain tissue, affecting the effectiveness of systemic chemotherapy. Localised drug delivery devices have emerged as a promising solution to address these limitations. This study focuses on developing and investigating biodegradable implantable devices to directly deliver chemotherapeutic drugs to brain tissue, reducing systemic administration and recurrence risk. Irinotecan (IRN) has shown success in clinical trials against GBM. However, large intravenous doses of IRN lead to severe systemic side effects. The study investigates the local delivery of IRN through implantable drug delivery devices to improve therapeutic outcomes while minimising adverse effects. The first study develops IRN-loaded poly(lactic-co-glycolic acid) (PLGA) implants using injection moulding (IM) and hot melt extrusion (HME) techniques. The study compared IM and HME implants, finding that HME implants showed better drug content uniformity and homogeneous drug distribution due to high shear mixing. Factors such as PLGA type, drug load, and implant size influenced drug release behaviour. HME implants exhibited slower drug release due to their denser matrix. Accelerated release studies showed both IM and HME implants had sustained release over seven days, with HME implants considered preferable based on drug content, stability, and distribution results. Pitavstatin (PTV) effectively slows tumour growth, but its limited BBB penetration suggests potential benefits of local administration. The second study involves the development of PTV-loaded PLGA implants using IM and HME techniques. Both IM and HME implants demonstrate an amorphous state of PTV. HME implants show higher drug content and uniformity due to homogeneous drug distribution facilitated by high shear mixing force, making them preferable over IM. In vitro drug release studies revealed slower drug release from HME implants due to denser matrices, and accelerated release studies confirmed sustained release over seven days for HME implants. In the third study, the high-performance liquid chromatography (HPLC) method was developed and validated for quantifying IRN and PTV in multi-layered implantable drug delivery devices. This method allowed precise drug release quantification, ensuring accurate safety and efficacy assessment of the devices. The final study involves the development of IRN-PTV-loaded PLGA implants as multi-layered devices using HME. The purpose of combining IRN and PTV in these implants is their synergistic effects against the GBM. The multi-layered implants were characterised using HPLC, differential scanning calorimetry (DSC), X-ray diffraction (XRD), and Raman spectroscopy. They demonstrated uniformity in size, weight, and drug content, validating the reliability of the HME technique. XRD and DSC analyses confirmed crystalline IRN in the IRN-PLGA layer and amorphous PTV in the PTV-PLGA layer, suggesting enhanced drug bioavailability and therapeutic effectiveness. Raman mapping reveals homogeneous drug distribution within the implants, ensuring consistent drug release. In vitro studies show biphasic drug release over seven days, characterised as non-Fickian behaviour by the Korsmeyer-Peppas model. This enhances our understanding of the release mechanism. This thesis presents advancements in implantable drug delivery devices for localised GBM treatment, offering valuable insights into formulation compositions and manufacturing techniques. Further research is necessary to assess in vivo performance and therapeutic efficacy
Exploring how antibiotics impair antifungal immunity
Get PDFProlonged antibiotic use is a controllable risk factor for systemic candidiasis. In recent years, antibiotics have been shown to have off-target detrimental effects on immune cells, but whether antibiotics disrupt antifungal immunity is unknown. Drummond and colleagues recently found that the antibiotic vancomycin enhanced mortality from invasive candidiasis in mice, whereas other antibiotics (e.g., metronidazole) did not affect susceptibility. To examine the mechanisms by which vancomycin impairs anti-Candida immunity, I explored how vancomycin pre-treatment of bone-marrow derived macrophages (BMDMs) impacted their responses to C. albicans challenge. Vancomycin-treated macrophages had reduced fungal killing, although phagocytosis of yeast cells was unaffected. Instead, vancomycin causes macrophage mitochondrial dysfunction which is accompanied with increased reactive oxygen species (ROS) production. ROS seemingly contributes to early inflammasome priming thus accelerating inflammasome activation and pyroptosis during C. albicans infections. Early inflammasome activation and pyroptosis potentially causes macrophages to have reduced time to exert all their antifungal properties on C. albicans thus causing reduced killing. Taken together, these results improve our understanding of the pathways regulating antifungal immunity in macrophages and suggest that antibiotic-induced susceptibility to C. albicans may be partly driven by disrupted macrophage function
Evaluation of 11β-hydroxysteroid dehydrogenase type 1 as a therapeutic target to treat sarcopenia in chronic kidney disease
Get PDFThe decline of muscle mass and strength, termed sarcopenia, is common in people with chronic kidney disease (CKD). Even though coexisting sarcopenia profoundly increases mortality and morbidity in CKD, no targeted treatments to correct the underlying disturbances in skeletal muscle metabolism are currently available.
This PhD thesis examines the hypothesis that excessive peripheral production of the hormone cortisol within skeletal muscle tissue by the enzyme 11beta hydroxysteroid dehydrogenase type 1 (11βHSD1) contributes to loss of muscle mass and strength in CKD, and that inhibition of 11βHSD1 has potential as a novel therapy to mitigate sarcopenia in CKD.
The scientific approach to explore this hypothesis comprises a range of preclinical and clinical research methodologies. The role of 11βHSD1 for muscle atrophy and consequences of its suppression are evaluated in vivo using a mouse model of persistent renal impairment. Complementary in vitro experiments using human muscle cells are conducted to elucidate regulation of 11βHSD1 by factors associated with CKD. Finally, 11βHSD1 activity is quantified directly in skeletal muscle biopsies in a case-control study involving volunteers with and without CKD to detect associations with kidney disease and muscle phenotype.
The data from independent preclinical and clinical lines of investigations indicate that 11βHSD1 activity in skeletal muscle tissue in conditions of CKD is unchanged compared to control conditions. Interestingly, the observational data from humans reveals that muscle 11βHSD1 activity is positively correlated with age and negatively correlated with muscle performance markers (grip strength, gait speed), regardless of CKD status. Nevertheless, genetic silencing of 11βHSD1 did not alter the muscle atrophy phenotype in mice with renal impairment. Compensatory elevation of circulating glucocorticoid levels is observed in mice without 11βHSD1 function, which may undermine the therapeutic efficacy of 11βHSD1 suppression.
In summary, the evidence gathered as part of this PhD thesis does not support significant upregulation of 11βHSD1 in skeletal muscle in response to reduction of kidney function or a major contribution to CKD-related sarcopenia risk. The upregulation of 11βHSD1 activity in skeletal muscle with age is consistent with previous reports and poses the question of its functional significance for age-related changes in skeletal muscle. The thesis concludes with a discussion of potential therapeutic niches for 11βHSD1 inhibitors that remain for future exploration
Change in the Quaker business environment 1800–1948
Get PDFThis thesis assesses change in the Quaker business community in Britain over the long period 1800–1948. It establishes a model for defining a Quaker business to assist in future research, and seeks to query notions of business success. I use a range of sources and methods and a largely chronological approach to draw out different strands of change over time. My earliest primary sources are the books of discipline of the Religious Society of Friends, which guided and regulated behaviour in its early history. By combining an analysis of the discipline around business with a consideration of Quakers’ business activities and the practice of the discipline, I show the impact on the Society and its reputation of Quaker interactions with new business forms and speculation in the early to mid-nineteenth century. I use a local case study of the Birmingham business community at this time to further demonstrate this, and to highlight the strength of Quaker business networks, locally and nationally, and how this trusting relationship benefits business.
My conviction that, based so much on personal faith, religious business history is also personal means that I highlight the cases of particular individuals, including Joseph Gibbins in the early nineteenth century, and the stories of two cousins and the family business they enter as young men in the late nineteenth century. These latter individuals, William Arthur Albright and John William Wilson, and the story of their firm and their differences draw out further loci of change. These are principally changes in company form and its associated importing of external management, and the devastating impact of World War I on the Quaker business environment. Throughout my thesis I bear in mind theological change in Quakerism and I argue that this change combines with the decline in the application of the discipline and leads to its replacement with a conversation which continues into the mid-twentieth century, but is ultimately inconclusive
Player perceptions of sports science monitoring practices in elite academy football
Get PDFIntroduction: Professional football players are subject to several different sports science and performance monitoring practices in an attempt to give applied practitioners a sound understanding of what training load coaches are prescribing to their players, but also to receive feedback on how players are reacting to the given training stimulus. To the writer’s knowledge however, little to no information exists in the current literature base that discusses what perceptions players themselves have of these monitoring practices. Study Purpose: The objectives of this study were to 1) Give athletes the opportunity to share experiences and feedback their own thoughts on monitoring methods, 2) Give applied practitioners feedback on how to successfully implement sports science monitoring practices from the opinion of the players themselves, 3) Provide evidence in an ongoing discussion surrounding the use of data in professional sport and inspire further study in the domain. Methods: 4 football players (age 17-21) from the development squad of a category 1 professional football academy took part in a 30-minute, semi-structured, focus group interview led by the head researcher to share their experiences of monitoring practices within their football academy. Thematic Analysis was then used to identify topics for wider discussion, and feed on recommendations to applied practitioners in the field and for further study. Results: Players appear to show a greater understanding and adherence to methods that are collected from them daily as opposed to practices that are more sporadic in nature. Player responses also highlighted the importance of player education regarding monitoring uses and the theory behind them in order to maximise adherence and limit negative perceptions of the practice or the practitioner. Conclusion: It is recommended that in order to help improve player perceptions of sports science monitoring practice that: 1) The monitoring methods are used as an educational tool for players. 2) This information is readily available to players should they wish to view it. 3) The information collected is used to inform practice or cause tangible change where possible. Indeed, these results reflect the reality of monitoring practices within one football academy and more research should be conducted with larger and more varied data sets to further investigate the perceptions of players more broadly in professional football.
Key Words: Monitoring, Football, Soccer, Subjective, Player Feedbac
Political economy of regionalism in Eurasia
Get PDFThis thesis addresses the question of why countries choose regional integration with some neighbouring states but not others. It contributes to the fields of comparative political economy and comparative regionalism. A theoretically informed empirical study examines two cases of the Central Asian (CA) republics of Kyrgyzstan and Kazakhstan and their respective integration decision with the Central Asian Economic Union (CAEU), the Eurasian Economic Union (EAEU) and the Shanghai Cooperation Organisation (SCO). The literature emphasises some factors over others as the main drivers of regional integration, namely economic benefits and costs, geopolitical concerns about external security, and the impact of domestic political systems. However, these insights do not satisfactorily explain why these CA republics, independent since 1991, still seek regional economic integration (REI) in Russia-led initiatives instead of integration within CA or wider Eurasia region. This study aims to fill this gap using the private and public interest framework. The latter of which includes consideration of economic and geopolitical components.
This thesis employs qualitative content analysis to generate findings and conclusions based on policy documents, policymakers' speeches, elite-level interviews, and trade and investment statistics. For the two case studies of these CA republics, the economic component dominated in the structure of the interests. The main determinants in the economic interests were existing trade and infrastructure dependencies and labour flows. The promise of economic modernisation and development was another factor favouring regional integration. Economic integration had to generate economic and external security benefits for these two CA republics. Political leaders favoured integration if non-participation risked their capacity to hold power and/or the integration initiative had the potential of generating an increased positive legacy of the incumbent’s time in office. The republics will, therefore, only choose integration if and when all three interests align
Manipulation of ultracold atomic gases with spatially and temporally controlled optical potentials
Get PDFThis thesis focusses on the use of a Digital Micromirror Device (DMD) to produce spatially and temporally varying optical potentials for the manipulation of ultracold atoms. I discuss the apparatus used to generate ultracold atomic samples as well as that used to manipulate them, including a comparison of the various light sources used with the DMD. Here I demonstrate the advantages in image quality that can be achieved using a superluminescent diode. I also report on work which uses the capability to produce spatially and temporally controlled potentials to engineer a synthetic dimension of harmonic oscillator sites, forming a lattice consisting of 10s of lattice sites on which Bloch Oscillations are demonstrated. Finally I describe simulations which demonstrate the clustering behaviour of quantised vortices in a Bose-Einstein condensate when arranged in linear structures, as could be done by the method of phase imprinting using a DMD. I demonstrate the use of a cluster detection algorithm which is novel in the field of quantised vortices, and show that in many cases, vortex clusters can be modelled by methods similar to those used for single vortices
Breaking out of the House of Silence: female voices and trauma in Arab women’s anglophone fiction
Get PDFThis thesis explores how female silence in Arab anglophone fiction functions paradoxically as a way of giving voice to women’s experiences, with speech and silence posited as reciprocal and constitutive parts of subjective expression. Five different novels by Arab women writers are examined and I argue that these utilise a similar range of narrative structures, framing strategies and themes that circumvent repression and marginalisation and allow for an alternative articulation of women’s dreams, desires and consciousness. The study finds a nuanced contrast between oppressive and communicative silences influenced by factors such as patriarchal power and gendered body politics, and explores how silence can become a means of introspection, self-preservation, resistance, observation or detachment.
I begin by locating the communicative function of silence within the frameworks of Bakhtinian polyphony, Diane Herndl’s conception of feminist dialogism and Arab women’s writing. The diversity and power of these women’s voices bring to the fore a heterogenous Arab female identity. The following chapters explore the shifting meanings and implications of female silence in the texts. Chapter one, ‘Embodying the Woman-Snake: Introspective Silence in Fatma: A Novel of Arabia’, explores Raja Alem’s portrayal of the transformative nature of women’s language and bodies and how silence is used to enact a spiritual journey that affirms female embodiment, empowerment and identity. The second chapter, ‘Women Behind Bars: Silence and Self-Preservation in The Golden Chariot’, examines how Salma Bakr’s ‘new language’ of feminine linguistic duality expresses women’s traumatic experiences, disrupting typical gender binaries and reclaiming agency through testimonies in the all-female space of a prison. ‘A Voice from Hell: Silence and Resistance in Woman at Point Zero’ investigates the relationship between female bodies, agency and silence in Nawal El Saadawi’s novel, allowing for the amplification and dissemination of female voices to challenge systematic abusive male authority. The contrast between dominant male narratives and female counter-narratives is explored in the fourth chapter, ‘Outsider, Madwoman, Exile: Narratives of Female Silence in Pillars of Salt’, as Fadia Faqir portrays the divergence of the male and female gaze and the creation of female counter-spaces. Chapter five, ‘Iron-Jawed Women: Hiding Behind Silence in Mornings in Jenin’, examines how, in Susan Abulhawa’s novel, silence becomes a protective defence mechanism in the face of individual and collective trauma, offering a new way of mapping the relationship between female voices and Palestinian identity in the face of displacement. Finally, the thesis ends with a conclusion that ties together the different threads of the chapters and considers the link between female voices, death and immortality
Learning to optimise reading: an exploratory study of eye movement transfer across languages
Get PDFThis dissertation provides an in-depth analysis of eye movement behaviour during reading in Chinese learners of English as a second language (L2, English), specifically focusing on the unique challenges posed by the divergent orthographic systems of the L1 and the L2.
At the core of this research lies a detailed exploration of how both language-general factors (e.g., word length and frequency) and task-specific factors (e.g., word order in the experiment) influence eye movement behaviours in reading across linguistically diverse systems. The significant orthographic disparity between Chinese and English provides a unique opportunity to examine how readers adapt their eye movements when navigating typologically distinct scripts, with a particular focus on controlling for common factors such as word frequency and length. Setting the eye movements of these L2 learners against eye movements from L1 reading in English and Chinese offers a broader perspective on reading behaviours across different linguistic backgrounds.
The thesis commences with a thorough literature review, contrasting the dense, spaceless appearance of the Chinese script with the more linear, spaced nature of English writing. This comparison lays the groundwork for a nuanced exploration into how readers modulate their eye movement patterns to accommodate the distinct linguistic structures and challenges presented by each script.
The eye movement investigation is structured around three central research questions. Firstly, it assesses how fixation durations and overall reading patterns vary among Chinese students reading in English, considering the dynamic evolution of these patterns in relation to varying levels of English proficiency and exposure. Secondly, the study investigates the influence of the Chinese script as an L1 on reading behaviours in English as an L2, examining the potential bidirectional transfer of reading strategies and the impact of prolonged exposure to English on reading behaviours in Chinese. Lastly, the research evaluates the facilitatory vs. inhibitory effects of prior exposure to similar content on reading processes in both L1 and L2, specifically probing how pre-reading similar content in either language under consideration affects subsequent eye movement patterns and reading strategies.
Employing advanced statistical techniques, including Principal Component Analysis (PCA), cluster analysis, and mixed effects modelling, this study provides a comprehensive understanding of the complex interplay between language proficiency, exposure, and cognitive mechanisms in reading across different scripts. The findings of this research significantly contribute to the field of cross-linguistic eye movement research and offer practical implications for language instruction and the development of reading interventions in multilingual contexts
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! 👇