Lublin University of Technology Journals
Not a member yet
4285 research outputs found
Sort by
Machine learning evidence towards eradication of malaria burden: A scoping review
Full text linkRecent advancements have shown that shallow and deep learning models achieve impressive performance accuracies of over 97% and 98%, respectively, in providing precise evidence for malaria control and diagnosis. This effectiveness highlights the importance of these models in enhancing our understanding of malaria management, which includes critical areas such as malaria control, diagnosis and the economic evaluation of the malaria burden. By leveraging predictive systems and models, significant opportunities for eradicating malaria, empowering informed decision-making and facilitating the development of effective policies could be established. However, as the global malaria burden is approximated at 95%, there is a pressing need for its eradication to facilitate the achievement of SDG targets related to good health and well-being. This paper presents a scoping review covering the years 2018 to 2024, utilizing the PRISMA-ScR protocol, with articles retrieved from three scholarly databases: Science Direct (9%), PubMed (41%), and Google Scholar (50%). After applying the exclusion and inclusion criteria, a final list of 61 articles was extracted for review. The results reveal a decline in research on shallow machine learning techniques for malaria control, while a steady increase in deep learning approaches has been noted, particularly as the volume and dimensionality of data continue to grow. In conclusion, there is a clear need to utilize machine learning algorithms through real-time data collection, model development, and deployment for evidence-based recommendations in effective malaria control and diagnosis. Future research directions should focus on standardized methodologies to effectively investigate both shallow and deep learning models
A concept for a production flow control system toolset for discrete manufacturing of mechanical products
Full text linkRequirements for product traceability in certain industrial sectors make Production Flow Control Systems (PFC) a desirable component in the operation of production enterprises. Such a system serves as a valuable tool for companies by preventing defective products from being sent to customers, enabling the automatic blocking of defective parts once the cause of the defect is identified. This article discusses a proposed PFC system toolset that meets fundamental industrial requirements in the field of discrete manufacturing of mechanical products. The system integrates key elements such as rework, disassembly, single non-controlled stations, as well as various essential and optional software applications and modules
From neglect to nurture: redefining pocket gardens as community vitality centers – case study of Al-Shorouk City, Egypt
No full textThis paper investigates the potential of pocket gardens to enhance community vitality, focusing on Al-Shorouk City in Egypt. Pocket gardens, often neglected, can be transformed into functional public spaces that foster community engagement and recreational use. This study employs a mixed-methods approach, including a literature review and observational studies to evaluate the physical attributes, maintenance, and usage patterns of three residential gardens in Al-Shorouk City. The findings reveal significant issues such as neglect, lack of amenities, and inadequate maintenance, which limit the gardens' functionality and social potential. Recommendations for improvement include regular upkeep, enhanced irrigation, diverse plantings, and the introduction of essential amenities. The study highlights the importance of community involvement and effective collaboration among stakeholders to create vibrant and sustainable urban green spaces. The insights gained can inform urban planning policies and practices to better integrate pocket gardens into the urban fabric, promoting environmental sustainability and social well-being
Informatyka i pomiary w opiece zdrowotnej: głębokie uczenie się w celu przewidywania ponownych hospitalizacji pacjentów z cukrzycą
No full textApproximately 460 million individuals were living with diabetes globally in 2023. This study explores and contrasts methods for forecasting hospital readmissions among diabetic patients by integrating traditional approaches with modern deep learning frameworks. In this work, a variety of deep learning architectures – including recurrent models like LSTM and GRU, as well as CNNs and Autoencoders – are examined along with conventional machine learning approaches. Four essential metrics – accuracy, precision, recall, and F1-score – were employed to measure and compare the effectiveness of different models. The results revealed that deep neural network methods significantly outperformed classical machine learning algorithms. Among traditional methods, the Decision Tree achieved the highest effectiveness. However, the LSTM network demonstrated superior performance, achieving scores of 0.74 for accuracy, 0.73 for precision, 0.74 for recall, and 0.73 for the F1-score. Additionally, the GRU and Vanilla LSTM models exhibited performance close to the best model, indicating that recurrent networks are more suitable for this problem than traditional methods.W 2023 r. na całym świecie na cukrzycę cierpiało około 460 milionów osób. Niniejszy artykuł analizuje i porównuje metody prognozowania ponownych hospitalizacji pacjentów z cukrzycą poprzez połączenie tradycyjnych podejść z nowoczesnymi frameworkami głębokiego uczenia się. W ramach niniejszej pracy przeanalizowano różne architektury głębokiego uczenia się – w tym modele rekurencyjne, takie jak LSTM i GRU, a także CNN i autoenkodery – wraz z konwencjonalnymi podejściami do uczenia maszynowego. Do pomiaru i porównania skuteczności różnych modeli wykorzystano cztery podstawowe wskaźniki – dokładność, precyzję, czułość i F1-score. Wyniki wykazały, że metody głębokich sieci neuronowych znacznie przewyższały klasyczne algorytmy uczenia maszynowego. Spośród metod tradycyjnych najwyższą skuteczność osiągnęło drzewo decyzyjne. Jednak sieć LSTM wykazała się lepszą wydajnością, osiągając wyniki 0,74 dla dokładności, 0,73 dla precyzji, 0,74 dla czułości i 0,73 dla F1-score. Ponadto modele GRU i Vanilla LSTM wykazały wydajność zbliżoną do najlepszego modelu, co wskazuje, że sieci rekurencyjne są bardziej odpowiednie dla tego problemu niż metody tradycyjne
Websites accessibility assessment of voivodeship cities in Poland
No full textThe aim of this study is to assess the accessibility of the websites of provincial capitals in Poland. The experiment consisted of an automated survey using five tools and an eye-tracking experiment with 15 participants. Analysis of the results showed that sites with elements without contrast errors are easier to locate, which translates into shorter time to first fixation, and the lowest time to first fixation was obtained by sites with tiled menus. In contrast, it is not possible to identify the best site from the results obtained. Each of the automatic tools evaluates the site according to its own established criteria. In contrast, the eye-tracking experiment carried out examined a small proportion of entire websites. Creating websites that comply with accessibility standards is key
Evaluating the effectiveness of selected tools in recognizing emotions from facial photos
No full textEmotion recognition from facial images has become a key area in computer vision and affective computing. Deep learning models such as convolutional neural networks and vision transformers have shown high potential in this domain. In this study, the performance of two representative architectures, ResNet-50, a convolutional neural networks based model, and ViT-B/16, a transformer-based model, is evaluated on the widely used Facial Expression Recognition 2013 dataset. Both models are trained using data augmentation and regularization techniques to enhance generalization. Their effectiveness is assessed using metrics including accuracy, precision, recall, and F1-score, alongside a detailed examination of confusion matrices. The observed differences in classification performance across emotion categories highlight the influence of architectural design on model behavior. The obtained results serve as a reference point for selecting appropriate deep learning architectures
SoundCrafter: Bridging text and Sound with a diffusion model
No full textText-to-sound systems have recently attracted interest for their ability to synthesize common sounds from textual descriptions. However, previous research on sound generation has shown limited generation quality and increased computational complexity. We present SoundCrafter, a text-to-sound generation framework that utilizes diffusion models. Unlike previous methods, SoundCrafter operates within a compressed domain of mel spectrograms and is driven by semantic embeddings derived from the CLAP model, which stands for contrastive language audio pretraining. SoundCrafter improves generation quality and computational efficiency by learning the sound signals without modeling the cross-modal interaction. In addition, we employ a curricular learning technique by progressively increasing spectrogram resolution to stabilize training and improve output fidelity. SoundCrafter distinguishes itself by integrating CLAP-conditional semantic embeddings with a diffusion model that operates in the compressed domain of mel-spectrograms. Using the AudioCaps dataset, it achieves superior text-to-sound synthesis with a Fréchet Distance (FD) of 23.45 and an Inception Score (IS) of 7.57 - exceeding the performance of previous models while requiring significantly less computational resources and training on a single GPU
Real-time detection of seat belt usage in overhead traffic surveillance using YOLOv7
No full textDriving safety plays a critical role in minimizing traffic accidents, and seat belt usage is one of the most effective preventive measures. This study aims to implement the YOLOv7 object detection model to automatically detect seat belt usage in four-wheeled vehicles using overhead traffic surveillance images. The proposed method consists of three main stages: dataset preparation, model training, and model evaluation. Dataset preparation includes acquiring video footage from different locations and time conditions, extracting image frames, and annotating four object classes: car, windshield, passenger, and seat belt. The model is trained on a dataset consisting of images taken during both day and night conditions. During training, data augmentation and anchor box optimization are applied to improve model generalization. The trained model is evaluated on an unseen test dataset and achieves a Mean Average Precision at 50% Intersection over Union (mAP50) of 97.46% and an F1 score of 95.37% at the optimal confidence level. These results indicate high detection accuracy for all object classes, especially for the seat belt class with an AP of 93.40%. The proposed system offers a promising solution for real-time traffic enforcement, reducing the reliance on manual observation and potentially improving traffic safety monitoring
Wodnosamoloty Macchi w Pucharze Schneidera, część 1.
No full textThe article describes the early starts of Macchi seaplanes in The Schneider Trophy, the annual race of seaplanes and flying boats.Artykuł opisuje początki startów wodnosamolotów Macchi w Pucharze Schneidera – corocznym wyścigu wodnosamolotów pływakowych i łodzi latających
Drewno jako materiał konstrukcyjny
No full textThe article concerns the use of wood as a material in aircraft construction. It presents the historical aspects of the use ofthis raw material in aircraft vehicles. The most commonly used types of wood are listed and characterised, and the basic technologies related to their use in aircraft construction are discussed. Examples of using woodin modern aircraft are also described.Artykuł dotyczy wykorzystania drewna jako materiału w konstrukcji statków powietrznych. Przedstawiono historyczne aspekty użyciatego surowca w budowie samolotów. Wymieniono i scharakteryzowano najczęściej stosowane gatunki drewna, a także omówiono podstawowe technologie związane z jego zastosowaniem w konstrukcjach lotniczych. Opisano także przykłady wykorzystania drewnawe współczesnych samolotach