Open Research Exeter - University of Exeter
Not a member yet
41213 research outputs found
Sort by
Hierarchical frequency-spatial collaborative method for extreme spatially-variant dehazing in remote sensing
Remote sensing (RS) images are susceptible to spatially nonuniform and irregular atmospheric haze caused by variations in atmospheric particle concentration, meteorological conditions, and topographic factors, which significantly degrades image visibility and contrast. This degradation significantly impairs applications in environmental monitoring, disaster assessment, and urban planning. The nonuniform distribution of atmospheric haze necessitates region-specific processing strategies. Additionally, conventional spatial-domain methods often fail to adequately recover high-frequency details lost due to atmospheric scattering. Although recent frequency-domain and deep learning-based approaches have demonstrated improvements in spectral recovery, they still struggle to simultaneously preserve structural details, visual quality, and spectral accuracy. These methods exhibit suboptimal performance when handling complex hazy scenarios. To address these challenges, we propose a frequency-aware hierarchical restoration network (FHR-Net) for extreme spatially variant dehazing in RS images. FHR-Net adopts a hierarchical frequency-aware restoration architecture for coarse-to-fine progressive restoration, effectively handling large-scale haze variations. Specifically, an adaptive multidomain feature reconstruction module utilizes a gating mechanism to dynamically balance spatial detail reconstruction and frequency-domain recovery, resulting in more precise feature extraction and detail preservation. Furthermore, we design a frequency-aware progressive multiscale loss (FAPM-loss) function that is highly compatible with the network’s multidomain processing characteristics, guiding the collaborative optimization of frequency-domain global recovery and spatial-domain detail reconstruction. Extensive experiments on synthetic RS datasets demonstrate that FHR-Net outperforms existing state-of-the-art (SOTA) methods, achieving improvements of 4.26% in PSNR over FFA-Net and 1.42% in SSIM over IRNeXt, thereby validating its effectiveness in complex and spatially variable dehazing scenarios.</p
Towards a Cultural Geography of Queer Storytelling through Collective Biography
This research project explores the power and value of queer storytelling through queer collective biographies – illustrated books containing collections of short biographies of LGBTQ+ people, telling the stories of queer lives from across the world, past and present, the famous and the everyday. This interdisciplinary research project brings together cultural geography, literary studies, and sexuality studies to establish a cultural geography of queer storytelling through collective biography which uncovers how these texts communicate knowledge, shape identities, create meaning and impact, and articulate cultural values.
Calling for literary geographies to pay attention to the genre of biography, this project uses a relational literary geographies approach, drawing on the conceptual frameworks of ‘text-as-spatial-event’ (Hones, 2008) and ‘book-as-assemblage’ (Anderson, 2015) to present a holistic understanding of ‘queer collective biography-as-assemblage’. This thesis asserts that the meaning and impact of queer collective biographies emerge out of an entangled network encompassing the interactions between an author-text-reader nexus, their relationships with external actors, and their intersections with spaces, and socio-political, cultural, and historical contexts. Methodologically, this thesis brings into conversation textual and visual analysis of eight queer collective biographies and semi-structured interviews with creators (authors, illustrators, editors, and publishers), circulators (teachers, queer bookshop owners, and librarians), and readers.
This project uncovers the complexities of writing queer lives, asking whose stories are told and by whom, through conceptualising ‘the queer publishing network’, shaped by concurrent personal motivations and commercial considerations. This thesis explores how queer lives are written, highlighting the challenges of navigating language and labels of queerness, privacy, and absence, when writing and illustrating queer lives, as well as uncovering the shortcomings of collective biography as a genre for telling queer stories. Moreover, this project draws attention to the narratives used to describe queer lives. Findings show that labels of heroes and icons perpetuate homonormativity; there must be critical engagement with the politics of visibility; and there is a need to platform queer joy whilst still attending to the difficult realities of LGBTQ+ life.
This thesis attends to the impact and meanings generated through readers’ encounters with books about queer lives through exploring the social, cultural and political possibilities (and limitations) of queer storytelling. It asserts that queer collective biography shapes understandings of identity and queerness in the world; provides affirmative, joyful, and empowering representation; creates a sense of community and belonging; promotes resistance and resilience; and provides an educational resource. Overall, this thesis argues that queer storytelling through collective biography is a powerful and valuable tool in times of crisis, fear, and hostility in the face of uncertain futures created by the anti-LGBTQ+ discourse that permeates UK society today.</p
<p></p><p dir="ltr">Meta-analysis of a decade of research on the accuracy of the Abbott Freestyle Libre continuous glucose monitoring system</p>
Objective: With a decade having passed since the introduction of the Abbott FreeStyle Libre continuous glucose monitoring (CGM) system in 2014, this study aims to conduct a comprehensive meta-analysis of its accuracy. We systematically reviewed and synthesized all published research evaluating the Libre system's performance, including studies comparing its glucose measurements to other devices, particularly the Dexcom CGM, and to clinical gold standards. Method: A meta-analysis was conducted, using studies published between 2014 and 2025. These studies were assessed across six categories: (1) Accuracy comparison between Libre and Dexcom CGMs, (2) Standalone accuracy assessments of Libre, (3) Clinical efficacy in diabetes management, (4) User-friendliness and patient experience, (5) Evaluation of trend arrows for direction-of-change insights, and (6) Libre performance during physical activity. The pooled data were analyzed using a random effects model to account for variability across studies. Result: Our meta-analysis of 15 datasets found no statistically significant difference in overall accuracy between the Libre and Dexcom systems. While 11 of the 15 studies reported a positive mean difference favoring Dexcom, the pooled mean difference was 0.96. However, the 95% confidence interval included the null value (0), indicating no significant advantage of one device over the other in terms of measurement accuracy when compared to reference standards. Conclusion: Accuracy of the Libre system can be affected by several factors: episodes of hypoglycemia, rapid glucose fluctuations, nighttime readings, exercise, sensor placement. Sensor performance may also vary on the first and last days of its 14-day wear period. Additionally, trend arrows indicating glucose direction of change warrant further scrutiny before being relied upon for decision-making. Notably, the Libre has shown acceptable accuracy in specific populations including children and pregnant women.</p
Filial piety and positive mental health: Investigating the roles of reciprocal and authoritarian filial piety in Singapore
The implications of filial piety on mental health and well-being have been increasingly examined within contemporary Eastern societies such as Taiwan, Hong Kong, and China. However, no studies to date have investigated the association between filial piety and positive mental health. The present study is the first to do so within the context of another East Asian society, Singapore. A total of 272 Singaporean participants (Mage = 30.0, SD = 7.7 years) completed an online survey assessing filial piety beliefs and positive mental health. Multiple hierarchical linear regression analyses were conducted to examine the extent to which filial piety beliefs accounted for variance in positive mental health outcomes. After controlling for potential confounders, both reciprocal and authoritarian filial piety positively predicted overall positive mental health. Specifically, reciprocal filial piety positively predicted the dimensions of general coping, emotional support, interpersonal skills, and personal growth and autonomy, while authoritarian filial piety was uniquely associated with spirituality. No significant associations were found between filial piety beliefs and global affect. These findings highlight the nuanced association between filial piety and broader conceptualizations of eudaimonic and hedonic well-being, contributing to the theoretical understanding of cultural beliefs and their role in mental health, and the subsequent development of culturally sensitive therapeutic interventions and policy-making.</p
Inclusive classrooms in Economics: understanding student engagement using Mentimeter
In this paper I examine whether different types of non-game-based audience polls affect students’ in-class performance and participation. I administered two types of Mentimeter polls in small group classroom setting for two first-year undergraduate Economics courses, one timed and the other un-timed quiz. Using ordinary least squares methods my analysis indicates that students seemed to perform 16-17% worse in timed quizzes. Interestingly, contrary to the existing literature, female students did not perform worse in these quizzes. Students were quite keen to participate in both types of quizzes with an overall average of approximately 80% of completion. Finally, students’ gender does not affect their participation or performance conditional on the gender of the tutorial leaders.</p
Antifungal Immune Responses and Inflammation in the Cystic Fibrosis Airways
Cystic Fibrosis (CF) is an inherited disease where the dysfunction of a single protein (the Cystic Fibrosis Transmembrane Regulator or CFTR), a chloride transporter found in both epithelial and immune cells, results in persistent infections and inflammation in the airways. It has been shown that Aspergillus fumigatus frequently infects the airways of people with CF and contributes to hyperinflammation and lung function decline.
The studies presented in this thesis aimed to obtain an insight into the airway mycobiome of people with CF, as well as getting a broader understanding of the host-fungal interactions. This was done by assessing the immune and inflammatory responses to a range of fungal pathogens by immune cells and epithelial cells.
My studies into the host-fungal interactions showed that the antifungal immune responses are fungus and fungal morphotype specific, and that the differences caused by the impaired function of the CFTR protein varies between host cell types. Aspergillus hyphae are more potent inducers of ROS than Aspergillus conidia, while Rasamsonia conidia are more potent inducers of ROS than Aspergillus conidia. No differences were observed in the antifungal responses between healthy and CF epithelial cells and neutrophils against Rasamsonia conidia. CF epithelial cells showed increased phagocytosis, but decreased killing, of Candida albicans blastospores. Interestingly, we observed that the previously described increase in Aspergillus-induced ROS production in CF neutrophils was no longer present in patients receiving treatment with the CFTR modulator Kaftrio. This strongly suggests that Kaftrio has an additional immunomodulating effect without compromising antifungal activity.
I explored if FRISA (Fungal Ribosomal Intergenic Spacer Analysis) could be a valuable and easy to perform technique to study the mycobiome in clinical samples. Unfortunately, due to shortcomings inherent with the technique, I was unable to obtain meaningful results.
As the CF antifungal immune responses are fungus and fungal morphotype specific and antifungal responses differ based on cell type and between healthy and CF cells, it is important to understand the composition of the mycobiome
3
within the CF airways to guide treatment. Especially during exacerbations, it is imperative to determine the best interventions to optimize clearance of the causative fungus and controlling the inflammation evoked. My results have shown that Kaftrio reduces the hyperinflammation against Aspergillus seen in the CF airways without effecting fungal clearance. And as a result of the widespread use of Kaftrio, and other CFTR modulators, the perspective of CF associated fungal disease may well be changing substantially.</p
Balancing autonomy and support in Ghana's irrigation governance
Water resource management in sub-Saharan Africa faces growing pressures from climate change, population growth, and rising competition over scarce water, especially in agricultural regions reliant on increasingly variable rainfall. While governments have long prioritized centralized management through formal irrigation schemes, informal farmer-led systems have become increasingly prominent. This study examines how institutional arrangements shape water governance by comparing water user associations (WUAs) in formal and informal irrigation schemes in Ghana's Upper East Region. Using Ostrom's design principles, and drawing on 160 interviews across 18 communities plus a participatory workshop, we analyse how different degrees of formalization influence infrastructure maintenance, community participation, and water allocation. Findings show that autonomy and external support are not opposing forces: support strengthens governance when it reinforces rather than replaces local decision-making. Infrastructure sustainability depends less on resource volumes than on alignment between governance structures and local maintenance capacities. Effective scaling requires retaining core elements of local autonomy while building targeted linkages to support institutions. We argue for moving beyond formal–informal dichotomies towards hybrid governance arrangements that combine state support with community-led management. These insights offer practical guidance for designing irrigation governance systems that enhance sustainability and equity in sub-Saharan Africa.</p
Long-term organizational growth following disasters: the role of collective empathy
This study examines how organizations exposed to disasters transform collective adversity into sustained growth. Drawing on organizational learning and collective emotion theories, we argue that disasters, while often disruptive in the short term, can also foster the emergence of collective empathy—an organizational-level sensitivity to others’ needs that motivates coordinated actions to support societal well-being—which, as organizations recover, contributes to long-term growth. Using difference-in-differences analyses on an 11-year panel of 575 Japanese companies affected by the 2011 Great East Japan Earthquake, combined with unique survey data, we find that organizations with moderate disaster exposure achieved, on average, higher long-term growth than those with low or high exposure. Further analyses suggest that this pattern is mediated by heightened collective empathy. The study identifies collective empathy as a key pathway to post-disaster organizational learning, underscoring the critical yet underexplored role of collective emotions in shaping organizational growth trajectories after disasters.</p
Who can see the men behind the curtain?
Review of Paranoid Finance, by Fabien Muniesa, Cambridge, Polity, 2024, 140 pp. ISBN 978-1-509-56116-2 (hardback); ISBN 978-1-509-56117-9 (paperback).</p
The hidden rise of toxicity: How retweets obscure increasing hostility in Brazilian politics
Toxicity on social media is often assessed using aggregate trends, yet different forms of engagement may shape these patterns in distinct ways. We analyse approximately 100 million Twitter posts collected during the 2018 Brazilian presidential election and the following year to examine how engagement type, automation, and user activity relate to online toxicity. Using a fine-tuned toxicity model, we find that 12% of posts are classified as toxic at a 0.5 threshold (6% at 0.8). Although aggregate toxicity appears relatively stable over time, this masks diverging engagement dynamics. Retweets, which account for about 60% of posts, are consistently less toxic (relative risk ≈ 0.5) and exhibit a slight downward trend. In contrast, replies are twice as likely to be toxic ( ≈ 2) and increase significantly over time. Replies to toxic parent posts are themselves more likely to be toxic ( ≈ 1.6), indicating conversational propagation of hostility. Contrary to the common assumption that automated ac counts primarily amplify content through retweeting, we find that higher automation levels are associated with fewer retweets and more replies and original tweets. Consistent with the higher toxicity of replies, automated accounts are disproportionately more likely to post toxic replies. By contrast, highly active accounts favour retweeting behaviour and, due to this amplification pattern, are not the primary drivers of toxicity. These findings demonstrate that rising hostility is concentrated in conversational exchanges and partially obscured by large volumes of low-toxicity amplification, challenging simplified assumptions about automation and online toxicity.</p