University of the West of Scotland
Research Repository and Portal - University of the West of ScotlandNot a member yet
10867 research outputs found
Sort by
Accelerated rotation-invariant convolution for UAV image segmentation
Full text linkRotation invariance is essential for precise object level segmentation in UAV aerial imagery, where targets can have arbitrary orientations and exhibit fine scale details. Conventional segmentation architectures like UNet rely on convolution operators that are not rotation-invariant, leading to degraded segmentation accuracy across varying viewpoints. Rotation invariance can be achieved by expanding the filter bank across multiple orientations; however, this significantly increases computational cost and memory requirement. In this paper, we introduce a GPU-optimized rotation-invariant convolution framework that eliminates the traditional data lowering (im2col) step required for matrix multiplication based convolution. By exploiting structured data sharing among symmetrically rotated filters, our method achieves multi-orientation convolution with greatly reduced memory requirements and computational redundancy. We further generalize the approach to accelerate convolution with arbitrary (non-symmetric) rotation angles. Integrated into a UNet segmentation model, the framework yields up to a 5.7% improvement in accuracy over the non-rotation-aware baseline. Across extensive benchmarks, the proposed convolution achieves 20–57% faster training and 15–45% lower energy consumption than cuDNN, while maintaining accuracy comparable to state of-the-art rotation-invariant methods. Because the scatter-based operator greatly reduces intermediate feature dimensionality, the efficiency of our design also enables practical sixteen-o rientation convolution and pooling, yielding further accuracy gains that areinfeasible for conventional rotation-invariant implementations. Our sixteen-orientation approach achieves competitive accuracy on multiple datasets compared with state-of-the-art UAV segmentation networks. These results demonstrate that the proposed method provides an effective and efficient alternative to existing rotation-invariant convolution frameworks
Managing through complexities:opportunities and challenges for rural female entrepreneurs in Bangladesh
Full text linkThis present chapter explores the contextual realities of rural female entrepreneurship in Bangladesh, a topic that remains underdeveloped in current research. It discusses the everyday challenges faced by women entrepreneurs and highlights the sociocultural and institutional dynamics that shape their business experiences. Using field data drawn from interviews with women operating in Dhaka, Rangpur and Khulna City, the study sheds light on how rural context influences entrepreneurial motivation, behaviour and outcomes. The authors identify a range of constraints including gender norms, financial exclusion and weak institutional support, which collectively impact the entrepreneurial journeys of these women. Central to the authors’ contribution is the recognition of women’s resilience and adaptive strategies in navigating these barriers. The study adds to the growing discourse on gender and entrepreneurship by presenting a more nuanced understanding of rural female entrepreneurship in South Asia
Preventing drug-related deaths in Scotland:perceptions and experiences of engagement in a “shared care” model of service delivery
Full text linkPurpose Scotland faces a crisis of drug-related deaths, disproportionately affecting people living in the most deprived areas. The purpose of this paper is to explore patient and service provider perceptions of engagement within shared care treatment systems, acknowledged as a critical factor in preventing drug-related harms and deaths.Design/methodology/approach A qualitative case study approach was adopted, focusing on two primary care practices in highly deprived urban areas. Thematic analysis was used to investigate the interplay of individual, organisational and structural factors acting as facilitators and barriers to service engagement. Data were collected through 34 semi-structured interviews with 6 people who use drugs, 4 family members, 20 health-care practitioners and 4 policymakers.Findings Engagement challenges were multifaceted, encompassing relational aspects (e.g. trust and stigma) and systemic issues, including poor collaboration across professional groups, fragmented services, inadequate communication and resource constraints. Participants emphasised the cumulative impact of socioeconomic deprivation and structural inequalities, which shaped the environments in which drug use occurred and constrained effective care delivery. Practitioners used various strategies, including harm reduction approaches and personalised support, to enhance engagement.Originality/value This paper provides new insights into the challenges faced by practitioners, people who use drugs and families in navigating the shared care system. The findings of this study highlight the need for policy action to strengthen service provision as well as reinforcing the importance of tackling cumulative health and social inequalities, seen as a key factor in drug-related deaths
Anomaly detection of multivariate finite-horizon process based on the rank-energy statistic
Full text linkIn this Industry 5.0 era, many online production processes resemble multivariate Finite Horizon Production (FHP) process, a time-bound stochastic process characterized by several variables. While the literature on monitoring multivariate FHP processes is scarce, these methods often rely heavily on the assumption of a multivariate normal distribution, which is challenging to achieve in real-world applications. To address this limitation, this article proposes a non-parametric exponentially weighted moving average (EWMA) control chart based on the rank-energy (RE2) statistic. The RE2 test, founded on the measure transportation theory, offers a robust approach for detecting shifts in multivariate process distributions. The robustness and anomaly detection ability of the proposed EWMA-RE2 chart is assessed using Monte Carlo simulations. Two real-world industrial production datasets are used to demonstrate the practical relevance for industrial applications. The proposed method displayed a stable performance for a reference sample of size more than 50 (m>50). Comparison study demonstrates a superior robustness and efficient shift detection ability of the proposed EWMA-RE2 chart over it's competitors, especially for multivariate processes with skewed distributions
Keeping trust when leaders go hybrid:a phase map and playbook for sustaining follower trust
Full text linkPurpose Hybrid leadership is increasingly recognized as an important feature of contemporary organizations, yet there is still no clear explanation of how hybrid settings change the evidence that followers use when judging a leader’s competence, fairness and care. This paper investigates how hybrid leadership changes the cues followers use to evaluate competence, fairness and care. In doing so, it provides an evidence-based framework that enables senior teams to sustain both cognition-based and affect-based trust through the intentional design of cue architecture.Design/methodology/approach The study adopts a structured narrative review in line with PRISMA 2020 guidance. Scopus was used as the primary database, identifying 216 records. After deduplication and screening, 70 empirical studies were retained. These were organized across four temporal phases : pre-crisis baseline, crisis, stabilization and maturation : and synthesized into four integrated pathways that leaders can implement to build and sustain trust.Findings The empirical evidence supports the proposition that hybrid leadership does not simply reduce proximity but rather recalibrates the distribution and interpretation of cues. The review highlights asymmetric vulnerabilities: affect-based trust shows resilience through rituals, care signals and symbolic gestures, while cognition-based trust erodes rapidly without visible proof of reliability and procedural fairness. Successful hybrid models deploy four mutually reinforcing mechanisms: predictable cadence that structures time, legibility of judgement through documentation, engineered proximity through designed touchpoints, and monitoring governance that creates visibility without surveillance. Phase progression from crisis to stabilization and then to maturation demonstrates organizational learning extending beyond short-term adaptation. Contextual factors such as task interdependence, tenure composition and equity concerns moderate effectiveness.Research limitations/implications The findings show that leaders must design predictable routines that make judgement visible and maintain human connection, particularly when teams rely on mediated interaction. Leaders should design predictable interaction points that balance focus with connection and govern monitoring through narrow, co-owned metrics. Trust requires quarterly review and rebalancing across both strands. The evidence base remains skewed toward knowledge-intensive sectors, and the automated screening method, although rigorous, carries classification risks.Originality/value This paper contributes by reframing hybrid leadership from location policy to cue architecture, offering a more nuanced account of how distance and justice interact with trust. It specifies the asymmetric mechanisms affecting cognition- and affect-based trust, integrates monitoring governance into the trust literature, and presents a phase-aware playbook that senior teams can apply. In doing so, it shows that sustaining trust in hybrid contexts requires intentional design rather than heroic availability
Association and agreement of contact-based smartphone photoplethysmography compared with electrocardiography:protocol for a systematic review and meta-analysis
Full text linkBackground: Mobile health (mHealth), leveraging mobile devices for health measurement and promotion, is rapidly growing. Smartphone cameras can perform photoplethysmography (PPG) to estimate pulse rate (PR) and other features of the cardiac cycle. However, establishing the validity of PR-PPG is essential before it can be adopted for health care applications. There is a pervasive belief that PR-PPG is analogous to heart rate (HR) derived from electrocardiograms (ECGs), and we will conduct a systematic review and meta-analysis to support or challenge this supposition.Objective: This study aims to synthesize quantitative evidence on the validity of PPG derived from mobile devices (ie, smartphones) for the assessment of HR compared with the gold standard ECG assessment.Methods: A comprehensive literature search will be performed on CINAHL Ultimate, MEDLINE, ScienceDirect, and Scopus using a predefined search strategy. All retrieved citations will be imported into Rayyan for screening and data management. A minimum of 2 independent reviewers will conduct the title and abstract screening, followed by 2 independent reviewers who will perform full-text screening and data extraction. All stages will be guided by predefined inclusion and exclusion criteria, which will be pilot-tested to ensure consistency and reliability. Any discrepancies will be resolved through discussion with a third reviewer or during a research team meeting. Intrarater reliability will be quantified at the title and abstract stage and the full-text review stage using Cohen κ. To ensure clarity and consistency in the presentation of study characteristics and findings, both narrative synthesis and tabular formats will be used. This review will include studies that report the association and agreement between resting HR and PR from PPG using contact-based smartphone devices versus ECG as the gold standard. PPG signals will be obtained using a contact-based approach, defined as finger-on-camera measurements with the smartphone’s built-in camera and flash. Studies will be excluded if they (1) do not use PPG using contact-based smartphone devices, (2) compare PPG to another collection method other than ECG, or (3) are review articles or case studies.Results: To inform clinical procedures and future studies, the results will contain data on PR-PPG and HR-ECG association (correlations) and agreement (Bland-Altman analysis), sampling devices, and operating systems. This project is unfunded, and the initial screening is expected to start in the first quarter of 2026, with results anticipated to be published in the first quarter of 2027. The projected timeline for the study includes title and abstract screening from the first quarter of 2026, followed by full-text screening in the second quarter of 2026. Results are anticipated by the third quarter of 2026, with publication expected in the first quarter of 2027. Throughout this period, database searches will be regularly updated to capture any newly published studies meeting the inclusion criteria.Conclusions: This review will provide a comprehensive understanding of the association and agreement between PR-PPG and HR-ECG. The findings may inform future adoption of PR-PPG and HR-ECG with insights into device or setting characteristics for best agreement or association.Trial Registration: Open Science Framework; https://doi.org/10.17605/OSF.IO/83V7AInternational Registered Report Identifier (IRRID):PRR1-10.2196/8483
Exploring agreement between golfers and coach in evaluating a putt
Full text linkThis study aimed to examine the previously unexplored area of agreement between golfers and a coach when assessing putts across two testing sessions. Eleven right-handed golfers (n = 6 males, 5 females, with the average putts per round 32.9 SD = 5.0) completed a 12-putt task, and a SAM PuttLab baseline at each session. The 12 putts varied by distance; short: 6–9 ft (1.83- 2.74m), medium: 12–15 ft (3.66–4.57m), long: 18–20 ft (5.49–6.10m), and slope (limited <1%, moderate >1% <2%, severe >2%). Following each of the 12 putts, the golfers and coach independently rated five variables: green reading, setup, pace, aim, and execution, using binary yes/no ratings. The coach had access to feedback via TrackMan. Cohen’s Kappa was used to assess agreement for each variable. Hierarchical multiple regression analyses examined predictors of putting performance, including hours of practice, SAM PuttLab consistency scores, and average putts per round. Results demonstrated low overall agreement across the two testing sessions. Regression analyses indicated that average putts per round was a significant predictor of performance at both testing sessions. Baseline kinematic consistency predicted performance at the first session, while practice hours did not predict performance at either session. These initial findings highlight the complexity of the perceptual-cognitive-motor skill of putting. Future research should continue to investigate how golfers and coaches develop shared affordances, using mixed method approaches, incorporating measures of expertise to gain greater understanding of how expert putting performance emerges and is maintained under variable environmental and individual conditions
Combined effects of nitrate and antimicrobial compounds on <i>in vitro</i> subgingival biofilms
Full text linkChlorhexidine and antibiotics are commonly used as adjunct treatments for periodontitis. However, these antimicrobials can lead to microbial resistance and chlorhexidine can impair health-associated nitrate (NO3-) metabolism. We tested the effect of chlorhexidine (0.002%), metronidazole (16 µg/ml) and amoxicillin (0.7 µg/ml), with and without 8 mM NO3-, on the bacterial composition and NO3- metabolism of subgingival plaque samples from 12 periodontitis patients grown in vitro for 8 h. The low sublethal concentrations of amoxicillin and chlorhexidine significantly inhibited microbial growth and impaired NO3- reduction, whereas the physiological concentration of metronidazole did not. A lower subgingival microbial dysbiosis index (SMDI) was found in the NO3- condition compared with amoxicillin alone and chlorhexidine with or without NO3- (p < 0.05). The SMDI of the metronidazole conditions was also significantly lower than in those with chlorhexidine (p < 0.05). Moreover, NO3- alone or combined with metronidazole appeared to increase Neisseria spp. and Aggregatibacter spp., whilst disease-associated changes were found in the chlorhexidine and amoxicillin conditions. Adding NO3- to metronidazole led to health-associated changes compared with metronidazole alone. In conclusion, low levels of amoxicillin and chlorhexidine limited microbial growth, impaired NO3- metabolism and were linked to disease-associated microbial profiles. A dual treatment of metronidazole + NO3- should be further investigated in clinical studies
Combined effects of nitrate and antimicrobial compounds on <i>in vitro</i> subgingival biofilms
Full text linkChlorhexidine and antibiotics are commonly used as adjunct treatments for periodontitis. However, these antimicrobials can lead to microbial resistance and chlorhexidine can impair health-associated nitrate (NO3-) metabolism. We tested the effect of chlorhexidine (0.002%), metronidazole (16 µg/ml) and amoxicillin (0.7 µg/ml), with and without 8 mM NO3-, on the bacterial composition and NO3- metabolism of subgingival plaque samples from 12 periodontitis patients grown in vitro for 8 h. The low sublethal concentrations of amoxicillin and chlorhexidine significantly inhibited microbial growth and impaired NO3- reduction, whereas the physiological concentration of metronidazole did not. A lower subgingival microbial dysbiosis index (SMDI) was found in the NO3- condition compared with amoxicillin alone and chlorhexidine with or without NO3- (p < 0.05). The SMDI of the metronidazole conditions was also significantly lower than in those with chlorhexidine (p < 0.05). Moreover, NO3- alone or combined with metronidazole appeared to increase Neisseria spp. and Aggregatibacter spp., whilst disease-associated changes were found in the chlorhexidine and amoxicillin conditions. Adding NO3- to metronidazole led to health-associated changes compared with metronidazole alone. In conclusion, low levels of amoxicillin and chlorhexidine limited microbial growth, impaired NO3- metabolism and were linked to disease-associated microbial profiles. A dual treatment of metronidazole + NO3- should be further investigated in clinical studies
The role of standardized technical packages in ensuring consistency and accuracy in apparel product development
Full text linkTech pack is a comprehensive communication document that provides detailed instructions and specifications for apparel production. It serves as contracts between design sections, production sections, and clients, ensuring consistency, accuracy, and effective communication throughout the product development process. This study aims to identify and examine the specific components of tech packs. Through surveys, interviews, and analysis of existing tech packs, the study validates the importance of each component and its contribution to the overall effectiveness of tech packs. The methodology involves hypothesis generation, questionnaire development, data collection through surveys, and data analysis using statistical techniques. Confirmatory factor analysis (CFA) technique was used to prove the model statistically by using AMOS and SPSS software. Results were improved by several iterations, and they verified that the obtained components of tech packs are necessary to get the right information from customers. Furthermore, this study will contribute to the improvement and standardization of tech pack development, ultimately enhancing productivity and ensuring consistent and accurate apparel product development