17628 research outputs found
Sort by
Translation and validation of Vietnamese version of the Birth Satisfaction Scale-Revised (BSS-R)
Get PDFBackgroundThe 10-item Birth Satisfaction Scale-Revised (BSS-R) is a quick and easy survey instrument recommended by the International Consortium for Health Outcome Measures as the tool of choice for measuring women’s birth satisfaction.AimTo translate and validate a Vietnamese-language version of the BSS-R.MethodA quantitative cross-sectional method was used to gather data post translation and back-translation of a Vietnamese version of the BSS-R (VN-BSS-R). Data collected were psychometrically evaluated using key indices of validity and reliability.ParticipantsVietnamese women who were within one month postpartum of birth (N = 383) took part in the study.ResultsFindings illustrate that a two-factor model offered excellent psychometric properties. With the two-factor VN-BSS-R, five items loaded onto a subscale ‘Positive birth experiences’ and the other five onto a second subscale ‘Negative birth experiences’. This two-factor model offered a fit to data (root mean square error of approximation [RMSEA] = 0.07, 90% confidence interval [CI] [0.05, 0.09], root square mean residual [RMSE] = 0.04 and comparative fit index [CFI] = 0.97). Mean scores for the exploratory factor analysis [EFA]-derived ‘positive’ and ‘negative’ sub-scales were 17.12 (SD 2.34) and 8.40 (SD 4.18) respectively.ConclusionThe translated and validated VN-BSS-R is a psychometrically robust tool for measuring birth satisfaction in Vietnamese postpartum women.The VN-BSS-R is available for use to measure experiences and perceptions of intrapartum care received by Vietnamese wome
Chester Brown
No full textBest known for his alternative comics, Chester Brown (b. 1960) is one of the most acclaimed and influential cartoonists of the last half century. This first biography provides a critical account of Brown’s life and career, highlighting his role in the evolving comics landscape and tracing his journey from self-publishing minicomics on the streets of Toronto to creating award-winning graphic novels.Characterized by often minimalist art and unconventional themes, comics such as Yummy Fur, Ed the Happy Clown, I Never Liked You, Louis Riel, and Paying for It have consistently pushed boundaries and confronted taboos. Chester Brown offers unique insight into Brown’s creative process as well the scope of his work and its larger cultural contexts. Organized chronologically, the book provides a full account of the artist’s career, beginning with his failed attempts to break into superhero comics and ending with discussions of his most recent work, in which he blends autobiography with political views on sex work and religion.The book also examines Brown’s extensive authorial revisions and considers how he has deployed both these and an increasingly voluminous amount of paratextual material in the service of creating a highly distinctive authorial persona that in turn cannot help but influence how we encounter and read his work. Chester Brown pulls back the curtain on this pioneering artist and emphasizes the inseparability of Brown’s art and life, including the myriad ways they have informed each other across the last four decades of comics history
Robust Real-time Audio-Visual Speech Enhancement based on DNN and GAN
Get PDFThe human auditory cortex contextually integrates audio-visual (AV) cues to better understand speech in a cocktail party situation. Recent studies have shown that AV speech enhancement (SE) models can significantly improve speech quality and intelligibility in low signal-to-noise ratios ( SNR < −5dB ) environments compared to audio-only (A-only) SE models. However, despite substantial research in the area of AV SE, development of real-time processing models that can generalise across various types of visual and acoustic noises remains a formidable technical challenge. This paper introduces a novel framework for low-latency, speaker-independent AV SE. The proposed framework is designed to generalise to visual and acoustic noises encountered in real world settings. In particular, a generative adversarial network (GAN) is proposed to address the issue of visual speech noise including poor lighting in real noisy environments. In addition, a novel real-time AV SE based on a deep neural network is proposed. The model leverages the enhanced visual speech from the GAN to deliver robust SE. The effectiveness of the proposed framework is evaluated on synthetic AV datasets using objective speech quality and intelligibility metrics. Furthermore, subjective listening tests are conducted using real noisy AV corpora. The results demonstrate that the proposed real-time AV SE framework improves the mean opinion score by 20% as compared to state-of-the-art SE approaches including recent DNN based AV SE models
Multi-scale integration with semantic embedding and adaptive excitation transformer for underwater optical image enhancement
No full textAs an important technology in the domains of ocean exploration and underwater robotics, underwater optical image enhancement has drawn significant attention. However, underwater images suffer from severe degradation due to wavelength-related optical attenuation, refraction, and scattering. The need to acquire high-quality underwater optical images poses challenges to current techniques. A dual-branch neural network based on multi-scale integration with semantic embedding and adaptive excitation transformer for underwater optical image enhancement is proposed in this paper. First, a semantic embedding network serving as the semantic branch is introduced to extract low- and high-level semantic features effectively to generate images with clearer edges and texture details. Concurrently, an enhancement branch containing adaptive excitation transformer is constructed to enhance local details while concentrating on global information. Since different feature channels correspond to different patterns of the input image, an adaptive excitation mechanism is deployed to achieve adaptive estimations of channel weights, highlighting the more representative channels while penalizing less important channels to enrich texture and eliminate blurring and color deviations. Additionally, a multi-scale dynamic integration module is designed. It establishes a correlation between the semantic and enhancement branches and adaptively selects prominent features to avoid over-enhancement and improve clarity and color deviation. Both the qualitative and quantitative results evaluated on public underwater optical image datasets show that our method outperforms the state-of-the-art methods in terms of subjective perception and evaluation metrics, indicating outstanding learning and generalization capabilities. Furthermore, excellent performance highlights the substantial benefits it contributes to downstream visual-related engineering tasks
A Hybrid Wasserstein GAN and Autoencoder Model for Robust Intrusion Detection in IoT
Get PDFThe emergence of Generative Adversarial Network (GAN) techniques has garnered significant attention from the research community for the development of Intrusion Detection Systems (IDS). However, conventional GAN-based IDS models face several challenges, including training instability, high computational costs, and system failures. To address these limitations, we propose a Hybrid Wasserstein GAN and Autoencoder Model (WGAN-AE) for intrusion detection. The proposed framework leverages the stability of WGAN and the feature extraction capabilities of the Autoencoder Model. The model was trained and evaluated using two recent benchmark datasets, 5GNIDD and IDSIoT2024. When trained on the 5GNIDD dataset, the model achieved an average area under the precision-recall curve is 99.8% using five-fold cross-validation and demonstrated a high detection accuracy of % when tested on independent test data. Additionally, the model is well-suited for deployment on resource-limited Internet-of-Things (IoT) devices due to its ability to detect attacks within microseconds and its small memory footprint of kB. Similarly, when trained on the IDSIoT2024 dataset, the model achieved an average PR-AUC of % and an attack detection accuracy of % on independent test data, with a memory requirement of kB. Extensive simulation results demonstrate that the proposed hybrid model effectively addresses the shortcomings of traditional GAN-based IDS approaches in terms of detection accuracy, computational efficiency, and applicability to real-world IoT environments
Perceptually Adaptive Multi-Microphone Audio System for Inclusive Group Communication and Environmental Awareness
Get PDFThis system proposes a perceptually grounded, user-behaviour-adaptive audio enhancement architecture that supports clarity, spatial stability, and inclusive listening in complex environments. Leveraging a triangulated microphone array (left ear, right ear, and smartphone), the system dynamically enhances relevant sound sources and suppresses unwanted noise based on the user’s head orientation, behavioural cues, and physiological signals. It is designed to function robustly across multiple real-world contexts, including noisy group conversations, high-motion activities, multilingual environments, and prosthetic auditory support
Partnership in the Classroom: Engaging Students Through Inclusive Student-Teacher Relationships to Advance Social Justice
Get PDFStudent-teacher relationships matter in creating inclusive student engagement opportunities in higher education. Student engagement is a wide-ranging topic, and much of the existing literature discusses student course representation (where student leaders gather feedback and work with staff to enhance the quality of courses) and curriculum co-creation (where students and teachers partner in decision-making regarding aspects of a course) as two distinct approaches. However, there is a notable paucity of prior empirical research comparing the ways that student representation and curriculum co-creation can lead to different forms of relationship-building. Therefore, this qualitative study first examined student course representatives’ and co-creators’ perceptions of effective student-teacher relationships in courses, followed by an analysis of the differences in those relationships between teachers and (a) course representatives and (b) curriculum co-creators. We identified five elements of effective student-teacher relationships within the classroom context that help students feel included, connected, respected, valued, and inspired. We found that different communication structures inherent in course representation and curriculum co-creation yielded distinct contributions and risks in building inclusive student-teacher relationships. To deepen understanding of inclusive student engagement, we explore opportunities to enhance these relationships and work towards meaningful partnerships between students and teachers that can advance social justice
Enhancing security in 6G-enabled wireless sensor networks for smart cities: a multi-deep learning intrusion detection approach
Get PDFIntroduction: Wireless Sensor Networks (WSNs) play a critical role in the development of sustainable and intelligent smart city infrastructures, enabling data-driven services such as smart mobility, environmental monitoring, and public safety. As these networks evolve under 6G connectivity frameworks, their increasing reliance on heterogeneous communication protocols and decentralized architectures exposes them to sophisticated cyber threats. To secure 6G-enabled WSNs, robust and efficient anomaly detection mechanisms are essential, especially for resource-constrained environments. Methods: This paper proposes and evaluates a multi-deep learning intrusion detection framework optimized to secure WSNs in 6G-driven smart cities. The model integrates a Transformer-based encoder, Convolutional Neural Networks (CNNs), and Variational Autoencoder-Long Short-Term Memory (VAE-LSTM) networks to enhance anomaly detection capabilities. This hybrid approach captures spatial, temporal, and contextual patterns in network traffic, improving detection accuracy against botnets, denial-of-service (DoS) attacks, and reconnaissance threats. Results and discussion: To validate the proposed framework, we employ the Kitsune and 5G-NIDD datasets, which provide intrusion detection scenarios relevant to IoT-based and non-IP traffic environments. Our model achieves an accuracy of 99.83% on the Kitsune and 99.27% on the 5G-NIDD dataset, demonstrating its effectiveness in identifying malicious activities in low-latency WSN infrastructures. By integrating advanced AI-driven security measures, this work contributes to the development of resilient and sustainable smart city ecosystems under future 6G paradigms
Optical and Electrical Properties of Amorphous Carbon Thin Films Grown from Mushroom Waste Oil using Chemical Vapor Deposition (CVD)
Get PDFAmorphous carbon thin films have been successfully deposited using the Chemical Vapor Deposition (CVD) technique with various amounts of natural oyster mushroom (P.ostreatus) waste oil as carbon precursors onto the glass substrates. This study examined the impact of varying precursor amounts on the assessment of the properties of amorphous carbon thin films, especially the optical and electrical characteristics. The lower part of the oyster mushroom, which is normally discarded, was harvested and extracted into mushroom waste oil using Soxhlet extraction. CVD technique was used for the deposition of amorphous carbon, where mushroom waste oil was used as a precursor and glass as a substrate. Then, UV-Vis spectroscopy, current-voltage (I-V) measurement, and Raman spectroscopy were utilized for characterization. The findings demonstrate that the thin films show a combination of sp3 and sp2 bonded carbon atoms, which is common for amorphous carbon. The lowest optical band gap, 0.37 eV, and the highest electrical conductivity, 1.51 × 10-5 S.cm-1 was deposited at 2 ml of oyster mushroom waste oil. These results indicate that amorphous carbon grown from mushroom waste oil is capable of being a ‘green’ alternative as a source for carbon-based solar cells in the future
Evaluation of Privacy-Preserving Support Vector Machine (SVM) Learning Using Homomorphic Encryption
Get PDFThe requirement for privacy-aware machine learning increases as we continue to use PII (personally identifiable information) within machine training. To overcome the existing privacy issues, we can apply fully homomorphic encryption (FHE) to encrypt data before they are fed into a machine learning model. This involves generating a homomorphic encryption key pair, where the public key encrypts the input data and the private key decrypts the output. However, there is often a performance hit when we use homomorphic encryption, so this paper evaluates the performance overhead of using an SVM (support vector machine) machine learning technique with the OpenFHE homomorphic encryption library. This uses Python and the scikit-learn library to create an SVM model, which can then be used with homomorphically encrypted data inputs and then produce a homomorphically encrypted result. The experiments include a range of variables, such as multiplication depth, scale size, first modulus size, security level, batch size, and ring dimension, along with two different SVM models, SVM-poly and SVM-linear. Overall, the results show that the two main parameters that affect performance are ring dimension and modulus size, and SVM-poly and SVM-linear show similar performance levels
Data sources:
Access Repository Dashboard
Do you manage Repository@Napier? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!