University of Glasgow Singapore

No abstract available.

Deepfake technology, which utilizes advanced artificial intelligence to create or manipulate multimedia content, presents significant challenges by obscuring the distinctions between reality and fiction. This phenomenon can lead to severe consequences such as misinformation and deception, particularly in culturally diverse regions like Southeast Asia. In response, this paper aims to enhance deepfake detection capabilities specifically for the Southeast Asian context, with a focus on Singapore, utilizing the Trusted Media Challenge (TMC) dataset. We employ transfer learning to fine-tune existing models with region-specific data and explore various layer freezing strategies to optimize performance. Additionally, we assess the effectiveness of transfer learning against the complete retraining of models to identify the most resource-efficient practices for improving detection capabilities. Our findings reveal that targeted fine-tuning of specific layers can enhance model performance in identifying regional deepfake, providing a balance between computational efficiency and detection accuracy. This research contributes to the development of robust, region-specific deepfake detection methods, which are crucial for combating the evolving threats posed by deepfake technology. We have developed a web application using our trained model, the web application and the source code are available at https://github.com/ict-at-sit/deepfake-detection-app.

Several studies have reported the prevalence of overweight and obesity in various countries but the global prevalence of nurses with overweight and obesity remains unclear. A consolidation of figures globally can help stakeholders worldwide improve workforce development and healthcare service delivery. To investigate the global prevalence of overweight and obesity among nurses. Systematic review with meta-analysis. 29 different countries across the WHO-classified geographical region. Nurses. Eight electronic databases were searched for articles published from inception to January 2023. Two independent reviewers performed the article screening, methodological appraisal and data extraction. Methodological appraisal was conducted using Newcastle-Ottawa Scale (NOS). Inter-rater agreement was measured using Cohen's Kappa. Meta-analyses were conducted to pool the effect sizes on overweight, obesity and waist circumference using random effects model and adjusted using generalised linear mixed models and Hartung-Knapp method. Logit transformation was employed to stabilise the prevalence variance. Subgroup analyses were performed based on methodological quality and geographical regions. Heterogeneity was assessed using the I2 statistic. Among 10,587 studies, 83 studies representing 158,775 nurses across 29 countries were included. Based on BMI, the global prevalence of overweight and obesity were 31.2% (n = 55, 95% CI: 29%-33.5%; p < .01) and 16.3% (n = 76, 95% CI: 13.7%-19.3%, p < .01), respectively. Subgroup analyses indicated that the highest prevalence of overweight was in Eastern Mediterranean (n = 9, 37.2%, 95% CI: 33.1%-41.4%) and that of obesity was in South-East Asia (n = 5, 26.4%, 95% CI: 5.3%-69.9%). NOS classification, NOS scores, sample size and the year of data collected were not significant moderators. This review indicated the global prevalence of overweight and obesity among nurses along with the differences between regions. Healthcare organisations and policymakers should appreciate this increased risk and improve working conditions and environments for nurses to better maintain their metabolic health. Not applicable as this is a systematic review. PROSPERO (ref: CRD42023403785) https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=403785. High prevalence of overweight and obesity among nurses worldwide.

Time series data is often processed to extract features which better explain the sources and structure of the data. However, these processes make underlying assumptions about the nature of the time series. Two important intrinsic properties are the linearity and stationarity of the data. The large corpora on time series analyses include domains of economics, physics and engineering – thus cross domain approaches can yield useful insights into the data. Here we look at data from accelerometers, an important class of sensors. We employ widely used time series tests to provide novel analyses to establish their linear and stationary structure. This provides useful insights into the underlying processes which are being sensed and guide the type of temporal features, any preprocessing needed and suitable analyses to be performed. We briefly mention the use of this in a machine learning application.

Accurate motion object detection (MOD) using in-vehicle cameras in driving vehicles is a challenging task. Several deep learning based motion segmentation approaches have been reported based on the interpretable optical flow feature. However, the interpretable optical flow feature has not been explored by object-level MOD approaches. In this paper, we propose a motion embedding pipeline (MEP) architecture that utilizes interpretable optical flow and deep learning to solve object-level MOD problems. The MEP is a three-stage pipeline that consists of an object detector, a novel feature extraction algorithm to capture relative motion between objects and the background, as well as a motion predictor for representation learning with stacked autoencoder to determine motions. A new dataset, Singapore motion object detection (SG-MOD) dataset is constructed in this work with much larger variations in urban environments. Experimental results show that the proposed MEP outperforms other pipeline-based architecture and deep learning based approaches on the SG-MOD and KITTI-MOD datasets in most metrics.

&lt;p dir="ltr"&gt;Autonomous driving systems (ADS) in recent years have been the subject of focus, evolving as one of the major mobility disruptors and being a potential candidate for deployment in urban cities due to urbanization. ADS is the system within the Autonomous Vehicle (AV) that enables automation. The different ADS technologies that enable autonomous vehicles have reached a certain maturity that no longer focus on technological deployment, but rather on the safe deployment on public roads. However, existing standards that validate functional safety and Risk Assessment (RA) may not be sufficient to tackle the increased complexity of ADS compared to traditional vehicles. This demand in ADS safety is exponentially increasing in tandem with the increase of AV automation levels. ADS are exposed to diverse environmental conditions and therefore subjected to operational risks while attempting to mimic the human driver responses. Moreover, the recent use of artificial intelligence and machine learning in the industry further shapes the way how ADS development will become in the future. This paper explains the importance of RA coverage for AV and provides a comparison and summary of existing RA methodologies. Thereafter, a recommendation of RAs for AV as potential solutions in meeting ISO 26262 and ISO/PAS 21448 standards.&lt;/p&gt;

During the engineering product development cycle, while going through the concept ideation stage, it is very important to accelerate the thinking process and to represent concept ideas in front of colleagues to brainstorm and refine for faster product design and its development. Availability and use of different computer-aided design (CAD) tools in the market made this concept generation process longer and time-consuming due to its limitation on detailing and editing. As an alternative, freehand sketching is very quick and easily editable which focuses on quantity rather than quality which allows healthy discussion and quick adjustment towards a valid and relevant design proposal. To explore learner's (students') experiences of concept ideation while studying product design and development, different freehand sketching tools were introduced like the use of some software instead of just traditional pen and paper kind of freehand sketching during the start of the trimester to quickly represent their ideas amongst their peers. During the later part of detail design, they used CAD tools. In this paper, a few examples are provided to show students' freehand sketching skills during concept ideation and the use of CAD tools for detailed design drawing. The effectiveness of this approach is measured through a survey conducted for students at the end of the trimester and it has been evaluated based on the survey results. They were asked to give responses based on their experience during the trimester. The results suggest that it is very important to emphasize the use of freehand sketching during concept ideation and the utility of CAD tools at the detail design stage of the product design development to achieve the learning outcomes of the module of Design and Manufacture 1 at the University of Glasgow, Singapore (UGS).

In higher education, software programming is an important fundamental skillset for students. Introductory software programming courses are offered to students majoring in not only computing science (CS), but also other schools including mechanical engineering, civil engineering, aerospace engineering, mathematics, business, and so on. Many students, particularly non-CS students, feel that software programming is not easy to learn and demonstrate good performance. Hands-on exercises in laboratories and tutorials are important to reinforce the learning of the theorical aspects of software programming. Clear guidance and immediate assistance from lecturers are highly appreciated in laboratory sessions to help students in their troubleshooting of software bugs. Teaching and learning face major challenges or even disruptions in the pandemic of COVID-19, due to the physical lock down of cities. Online learning or home-based learning has become a norm during the pandemic. As such, it introduces additional difficulty in the conduct of hands-on programming laboratory sessions, as lecturers and students are not in the same physical classroom. It may take a longer time for students to troubleshoot small software bugs in a programming code without immediate advice from lecturers. Students may become anxious or de-motivated in hands-on programming. In this paper, an online collaborative programming platform is employed to conduct laboratory sessions in the learning of Python programming subject, which is offered to Year 1 Civil Engineering students. It enables students and lecturers to work together or troubleshoot programming codes “virtually” from different physical locations. A case study and its effectiveness are explored by comparing the results before and after utilizing the collaborative programming. Interesting results are discussed on the learning process of Civil Engineering students who do not have any prior Python programming skills.