Evaluation Of Architecture Research Articles

Advancements in Target Detection: A Comparative Analysis of Deep Learning Models for Real-Time Applications

Abstract. Target detection is a vital field within computer vision, playing an essential role in applications. This paper investigates the advancements and efficiencies of contemporary target detection methodologies, focusing on deep learning frameworks. Through a systematic review and evaluation of these models' architectures and performances using the Common Objects in Context (COCO) dataset, the study highlights their operational effectiveness and practical implications in real-world scenarios. A detailed comparative analysis is conducted, assessing the models based on mean Average Precision (mAP) and input dimensions to determine their suitability across various detection tasks. You Only Look Once version 3 (YOLOv3), in particular, is recognized for its ability to combine high-speed detection with significant accuracy, effectively addressing real-time processing challenges. The results confirm that YOLOv3, despite its smaller input size, performs comparably to more complex systems, demonstrating notable enhancements in design efficiency and processing speed. This research underscores the potential for future optimizations in model architectures to bridge the gap between high-speed and high-accuracy detection tasks, potentially transforming real-time applications across multiple sectors. The practical significance of this work lies in its ability to guide future developments in object detection, benefiting both academic research and industrial applications.

Non-Destructive Estimation of Paper Fiber Using Macro Images: A Comparative Evaluation of Network Architectures and Patch Sizes for Patch-Based Classification

Over the years, research in the field of cultural heritage preservation and document analysis has exponentially grown. In this study, we propose an advanced approach for non-destructive estimation of paper fibers using macro images. Expanding on studies that implemented EfficientNet-B0, we explore the effectiveness of six other deep learning networks, including DenseNet-201, DarkNet-53, Inception-v3, Xception, Inception-ResNet-v2, and NASNet-Large, in conjunction with enlarged patch sizes. We experimentally classified three types of paper fibers, namely, kozo, mitsumata, and gampi. During the experiments, patch sizes of 500, 750, and 1000 pixels were evaluated and their impact on classification accuracy was analyzed. The experiments demonstrated that Inception-ResNet-v2 with 1000-pixel patches achieved the highest patch classification accuracy of 82.7%, whereas Xception with 750-pixel patches exhibited the best macro-image-based fiber estimation performance at 84.9%. Additionally, we assessed the efficacy of the method for images containing text, observing consistent improvements in the case of larger patch sizes. However, limitations exist in background patch availability for text-heavy images. This comprehensive evaluation of network architectures and patch sizes can significantly advance the field of non-destructive paper analysis, offering valuable insights into future developments in historical document examination and conservation science.

Kidney Disease Classification

Kidney diseases, such as tumors, cysts, and stones, are serious health conditions that affect millions globally. Medical imaging, particularly Computed-Tomography (CT) scans, plays an important role in diagnosing these conditions. Although, manual analysis of CT images by radiologist can be tiresome, fallible, and affected by human variability. This paper introduces a deep learning approach that employs Convolutional Neural Networks (CNN) for the automated classification of kidney CT scan images into four distinct categories : normal, cysts, tumors, and stones. The system is designed to boost diagnostic accuracy, reduce human error, and expedite clinical decision making. Through data preprocessing, CNN architecture design, training, and evaluation, the proposed model achieved a classification accuracy of 92%. This study highlights the significance of CNNs in medical image analysis anddtheir implicit for real- time deployment in clinical surroundings. Additionally, we compare our CNN-based approach with other advanced architectures like Vision Transformers, demonstrating CNNs' robustness and efficiency in this domain.

FedNIC: enhancing privacy-preserving federated learning via homomorphic encryption offload on SmartNIC

Federated learning (FL) has emerged as a promising paradigm for secure distributed machine learning model training across multiple clients or devices, enabling model training without having to share data across the clients. However, recent studies revealed that FL could be vulnerable to data leakage and reconstruction attacks even if the data itself are never shared with another client. Thus, to resolve such vulnerability and improve the privacy of all clients, a class of techniques, called privacy-preserving FL, incorporates encryption techniques, such as homomorphic encryption (HE), to encrypt and fully protect model information from being exposed to other parties. A downside to this approach is that encryption schemes like HE are very compute-intensive, often causing inefficient and excessive use of client CPU resources that can be used for other uses. To alleviate this issue, this study introduces a novel approach by leveraging smart network interface cards (SmartNICs) to offload compute-intensive HE operations of privacy-preserving FL. By employing SmartNICs as hardware accelerators, we enable efficient computation of HE while saving CPU cycles and other server resources for more critical tasks. In addition, by offloading encryption from the host to another device, the details of encryption remain secure even if the host is compromised, ultimately improving the security of the entire FL system. Given such benefits, this paper presents an FL system named FedNIC that implements the above approach, with an in-depth description of the architecture, implementation, and performance evaluations. Our experimental results demonstrate a more secure FL system with no loss in model accuracy and up to 25% in reduced host CPU cycle, but with a roughly 46% increase in total training time, showing the feasibility and tradeoffs of utilizing SmartNICs as an encryption offload device in federated learning scenarios. Finally, we illustrate promising future study and potential optimizations for a more secure and privacy-preserving federated learning system.

Mathematical Evaluation of Deep Learning Architecture with Feature Fusion for Cervical Cancer Detection Classification

Cervical cancer is still one of the main reasons women die around the world, which shows how important it is to have accurate and quick ways to diagnose the disease. This study looks into how to improve the accuracy of finding cervical cancer using a deep learning-based classification system that combines features. In this mixed classic feature extraction methods, like Histogram of Oriented Gradients (HOG) and Scale-Invariant Feature Transform (SIFT), with deep learning architectures, like Graph Convolutional Networks (GCNs) and Capsule Networks. Each of these added its own benefits to the fusion model. Both HOG and SIFT are good at capturing important texture and structure features, even when the image conditions change. GCNs are used to learn relationship data between picture areas. This helps get a better sense of how things fit together in space, which is very important for medical imaging. Capsule Networks, on the other hand, let you understand the subtleties of hierarchical patterns in pictures of cervical cells. They do this by looking at spatial hierarchies and rotational changes that are very important for finding cancer. Our suggested fusion model combines these multi-level features, taking advantage of how the best features of deep learning and standard learning methods work together. Validation on standard cervical cancer image datasets showed that this fusion approach was much more accurate and precise at classifying than individual methods. Quantitative testing showed that the fusion model had higher sensitivity and specificity, which shows that it can reduce the number of fake negatives and positives, which is very important in clinical diagnosis. Our results show that the feature fusion model not only gives us a strong and accurate way to classify cervical cancer, but it also lays the groundwork for more uses in other complicated areas of medical imaging.

Comparative Evaluation of RNN Architectures for Churn Prediction in Telecom Services

Comparative Evaluation of RNN Architectures for Churn Prediction in Telecom Services

Brachial and Lumbosacral Plexopathies.

The brachial and lumbosacral plexuses are complex neural structures that transmit sensory, motor, and autonomic information between the spinal cord and the extremities. Plexus disorders can be particularly disabling because lesions in the plexus usually affect large groups of nerve fibers originating from several spinal levels. Electrodiagnostic studies are often required to confirm a plexus lesion and determine the extent of injury and prognosis. Magnetic resonance is the imaging modality of choice for detecting intrinsic nerve abnormalities; recently, high-resolution ultrasound has emerged as an alternative method for dynamic evaluation and visualization of internal nerve architecture. Once a plexopathy is confirmed, the list of possible etiologies is relatively limited and includes traumatic and nontraumatic causes. Treatment relies on symptom management and physical rehabilitation unless a treatable underlying condition is found. Surgical approaches, including nerve grafts or tendon transfers, may improve limb function when spontaneous recovery is suboptimal.

Investigation, Evaluation, and Dynamic Monitoring of Traditional Chinese Village Buildings Based on Unmanned Aerial Vehicle Images and Deep Learning Methods

The investigation, evaluation, and dynamic monitoring of traditional village buildings are crucial for the protection and inheritance of their architectural styles. This study takes traditional villages in Shandong Province, China, as an example, employing UAV images and deep learning technology. Utilizing the YOLOv8 instance segmentation model, it introduces three key features reflecting the condition of traditional village buildings: roof status, roof form, and courtyard vegetation coverage. By extracting feature data on the condition of traditional village buildings and constructing a transition matrix for building condition changes, combined with corresponding manual judgment assistance, the study classifies, counts, and visually outputs the conditions and changes of buildings. This approach enables the investigation, evaluation, and dynamic monitoring of traditional village buildings. The results show that deep learning technology significantly enhances the efficiency and accuracy of traditional village architectural investigation and evaluations, and it performs well in dynamic monitoring of building condition changes. The “UAV image + deep learning” technical system, with its simplicity, accuracy, efficiency, and low cost, can provide further data and technical support for the planning, protection supervision, and development strategy formulation of traditional Chinese villages.

Optimizing Touchless Fingerprint Identification: A Machine Learning Approach to Modelling and Performance Evaluation

This paper explored the modelling and performance analysis of a smartphone-based fingerprint identification system using Convolutional Neural Networks (CNN). The research developed a theoretical framework to validate picture-based fingerprint identification as a feasible alternative to traditional touch-based methods. A modified Automated Fingerprint Identification System (AFIS) model served as the study's foundation. To enhance the model's capabilities, data from two databases, IIT India and SOCOFing, were utilized. The evaluation of the CNN architecture focused on mobile device fingerprint recognition. It emphasized key processes such as data pre-processing, model training, and the optimization of the CNN through a Siamese-CNN approach to boost accuracy and efficiency. Python scripts developed for this purpose were converted to Android code using TensorFlow for deployment on Android devices. Performance metrics, including identification accuracy, processing speed, and resource utilization, were analysed to determine the system's feasibility. The results demonstrated that CNN-based fingerprint identification systems hold significant promise for delivering robust and reliable biometric authentication on smartphones, highlighting both their practical applications and limitations. decrease medical as well as financial burden, hence improving the management of cirrhotic patients. These predictors, however, need further work to validate reliability.

Criteria for evaluation of modern architecture in Bosnia and Herzegovina to support its preservation

PurposeAs detailed criteria for evaluation of modern architecture do not yet exist in Bosnia and Herzegovina (B&H), the aim of this paper is to assist in their development by providing an analysis of the scope of those criteria in the related international documents. The research objective is to identify what values of modern architecture were historically and are currently considered significant to merit its preservation.Design/methodology/approachThe research methods include a review of relevant literature, collecting data from the B&H and international organisations criteria for evaluation of architectural heritage; a comparative analysis of the collected data to propose criteria for evaluation of modern architecture in B&H that are aligned with the evaluation criteria of UNESCO’s World Heritage List and with Docomomo’s criteria.FindingsThe above comparative analysis showed that such criteria could be developed by selecting relevant criteria from a long list of criteria for evaluation of built heritage in B&H, which currently do not consider architecture built after 1960.Practical implicationsThe proposed evaluation criteria will be tested on different typologies of modern architecture in B&H to enable wider discussion of stakeholders and relevant agencies on their applicability.Social implicationsThe proposed evaluation criteria will contribute to the promotion of architectural heritage of the 20th century in B&H.Originality/valueThere has been no previous research on how to develop criteria for evaluation of architectural heritage of the 20th century in B&H.

Evolving Adversarial Training (EAT) for AI-Powered Intrusion Detection Systems (IDS)

Intrusion Detection Systems (IDS) are crucial components of network security, yet traditional IDS models often fail to cope with rapidly evolving adversarial attacks that exploit their static nature. This study proposes a novel approach, Evolving Adversarial Training (EAT), to enhance the adaptability and robustness of AI-powered IDS against dynamic threats. The EAT framework integrates continuous model evolution with advanced adversarial training techniques, enabling the IDS to dynamically adjust to new attack patterns. Experimental results demonstrate that the EAT framework significantly enhances IDS performance, leading to increased detection accuracy and reduced false positive rates compared to conventional methods. These findings emphasize the potential of EAT in fortifying network defenses against evolving cyber threats, offering a promising avenue for future research in scalable and adaptive IDS solutions that can effectively combat the complexities of modern cyber adversaries. The research explores three key objectives: dynamic adaptation and adversarial training, continuous learning and enhanced threat detection, and robustness and generalization. By focusing on these objectives, the study aims to develop AI-powered IDS that can effectively navigate the ever-changing cyber threat landscape. The research methodology includes data collection, model architecture design, training and evaluation, continuous learning, simulation, and real-world testing, all aimed at enhancing the resilience of AI-powered IDS against adversarial attacks. By systematically following this framework, the study intends to enhance the security system of IDS through the effective implementation of EAT.

Field effect transistor biosensors for healthcare monitoring

AbstractThe burgeoning demand for healthcare monitoring has brought field effect transistor (FET) biosensors into the spotlight as a highly efficient detection technology. FET biosensors offer inherent advantages including high sensitivity, rapid response times, operational simplicity, integrability, and label‐free detection. These characteristics render them particularly well‐suited for detecting a diverse array of physiological parameters and biomarkers, thereby furnishing real‐time data crucial for personalized medicine and disease prevention. This review aims to elucidate recent advancements in FET biosensors within the realm of healthcare management encompassing several facets. Initially, this review systematically analyzes the device architecture, sensing mechanisms, and performance evaluation methods of FET biosensors to gain an in‐depth understanding of their operational principles and features. Subsequently, it focuses on the application of FET biosensors for detecting health‐related biomarkers encompassing nucleic acids, proteins, exosomes, viruses, etc. Lastly, the review presents engineered medical sensor prototypes predicated on FET biosensors, such as point‐of‐care testing devices, wearable sensors, and implantable sensors, underscoring their practical utility and potential in health management. In addition, the review addresses critical issues and prospects of using FET biosensors in healthcare monitoring, aiming to provide some references and insights for research and innovation in this field.

Architectural Evaluation of Eco-Industrial Parks within the Scope of Sustainable Human Development Goals

This study makes an architectural evaluation of eco-industrial parks within the scope of sustainable human development goals. Eco-industrial parks are areas of critical importance in reducing the environmental impacts of industrial activities. These parks offer innovative and effective solutions to reduce industrial pollution and greenhouse gas emissions. In this direction, the study evaluates the performance of eco-industrial parks according to criteria such as energy efficiency, material use, water management, waste management, indoor quality, landscape-green areas, social participation and economic sustainability. The sample structures of the research are Kokkola Eco-Industrial Park, Kalundborg Eco-Industrial Park, The Ecofactorij, Tianjin Economic Development Zone, Eco3 Industrial Park and Daedeok Technovalley Park, which are pioneers in the field of eco-industrial parks in the world. The findings of the study show that taking architectural criteria into consideration within the scope of sustainable development goals in the design of eco-industrial parks leads to more efficient results. Although pioneering eco-industrial parks have positive features in many aspects, they are inadequate in terms of interior quality, transportation, accessibility and innovative design. On the other hand, material selection, water and waste management are the criteria in which parks are most successful. According to the research findings, the parks with the highest scores were Kalundborg and TEDA, while the parks with the lowest scores were Ecofactorij and Kokkola. As a result, this study provides guidance to architects, urban planners and politicians on improving existing parks and shaping new designs. In the future, more comprehensive studies can be conducted on the design and planning of eco-industrial parks from the perspectives of different disciplines.

Evaluation of deep neural network architectures for authorship obfuscation of Portuguese texts

Preserving authorship anonymity is paramount to protect activists, freedom of expression, and critical journalism. Although there are several mechanisms to provide anonymity on the Internet, one can still identify anonymous authors through their writing style. With the advances in neural network and natural language processing research, the success of a classifier when identifying the author of a text is growing. On the other hand, new approaches that use recurrent neural networks for automatic generation of obfuscated texts have also arisen to fight anonymity adversaries. In this work, we evaluate two approaches that use neural networks to generate obfuscated texts. The first approach uses Generative Adversarial Networks to train an encoder–decoder to transform sentences from an input style into a target style. The second one trains an auto encoder with Gradient Reversal Layer to learn invariant representations. In our experiments, we compared the efficiency of both techniques when removing the stylistic attributes of a text and preserving its original semantics. Our evaluation on real texts clarifies each technique’s trade-offs for Portuguese texts and provides guidance on practical deployment.

Evaluation of Business-Driven Reference Architecture for Big Data Analytics Implementation by Public Sector Organizations in Resource-Constrained Setting: A Case Study of Uganda

Big Data Analytics (BDA) is a new area at the nexus of agenda, public sector organizations, and government business. It may satisfy the growing need for trustworthy, cost-effective services in the public sector for better, more informed decision-making processes. BDA has been proposed on the planning schedules of several public sector organizations and the government. Therefore, from the previous work, using Uganda as a case study, specifically the Uganda Bureau of Statistics (UBOS), Ministry of Health (MoH), and Ministry of Education and Sports (MoES), this study aims to evaluate a designed Business-Driven Reference Architecture for Big Data Analytics Implementation (BRABDAI) in public sector organizations. Using a quantitative research method, the BRABDAI was analytically assessed using a structured walkthrough technique by 18 IT experts in the fields of data analysis, IT/IS management, and solution architecting. The assessment criterion was composed of three components: functionality, usability/applicability, and traceability. The Statistical Package for the Social Sciences (SPSS) software was used for data analysis. The findings showed that the BRABDAI was useful for resolving the key issues preventing the successful implementation of the BDA in public sector organization settings and that it was also understandable and traceable. To improve the usage and acceptability of BDA implementation in public sector organizations, the study advises more research on this topic.

Design and Implementation of an Employee Health Management Information System

This study developed an information system for employee health management aimed at enhancing workplace health and increasing corporate competitiveness. The employee health management information system (EHMIS) integrates health data analytics, risk assessments, and personalized health intervention measures to promote a healthier workforce. The platform emphasizes data privacy, seamless integration with existing human resources systems, and provides tools for monitoring health trends, reducing absenteeism, and improving overall productivity. The building process includes needs assessment, system architecture design, information security, functionality integration, and evaluation. The implementation strategy prioritizes employee engagement and continuous improvement, effectively assisting companies in human resource management and ongoing health promotion for employees.

Design and Performance Evaluation of a Novel High-Speed Hardware Architecture for Keccak Crypto Coprocessor

Design and Performance Evaluation of a Novel High-Speed Hardware Architecture for Keccak Crypto Coprocessor

Developing a Training Program for a Personal Navigation Intervention for Individuals with Disabilities

Purpose: The purpose of this project was to identify, develop, and assess the feasibility of a complex intervention to support independent public transportation for individuals with cognitive disabilities in a midsized urban setting. Methods: We used the Medical Research Council framework to develop and evaluate a complex intervention. The complex intervention was an education and training program that supported personal navigation for individuals with disabilities (PNID) and was part of the Smart Columbus Mobility Assistance for People with Cognitive Disabilities (MAPCD) project. The identification, development, and feasibility of the PNID education and training program was iterative and included seven stages, which started with engaging with external stakeholders and ended with describing future directions for the implementation of the complex intervention. Results: The result was an evidence-based, theory-informed, and stakeholder supported education and training program to support public transportation for individuals with cognitive disabilities. The intervention was based on frameworks, models, and theories including a sociotechnical architecture model. The sociotechnical architecture was operationalized using the WayFinder system, a personal transportation assistant that consists of both a smartphone application and a web-based portal. The education and training program consisted of five activities. Conclusions: The PNID education and training program and MAPCD project provide a framework for the identification, development, implementation, and evaluation of sociotechnical architectures that support independent public transportation for individuals with disabilities in midsize urban settings.

Multimodal representations of biomedical knowledge from limited training whole slide images and reports using deep learning

The increasing availability of biomedical data creates valuable resources for developing new deep learning algorithms to support experts, especially in domains where collecting large volumes of annotated data is not trivial. Biomedical data include several modalities containing complementary information, such as medical images and reports: images are often large and encode low-level information, while reports include a summarized high-level description of the findings identified within data and often only concerning a small part of the image. However, only a few methods allow to effectively link the visual content of images with the textual content of reports, preventing medical specialists from properly benefitting from the recent opportunities offered by deep learning models. This paper introduces a multimodal architecture creating a robust biomedical data representation encoding fine-grained text representations within image embeddings. The architecture aims to tackle data scarcity (combining supervised and self-supervised learning) and to create multimodal biomedical ontologies. The architecture is trained on over 6,000 colon whole slide Images (WSI), paired with the corresponding report, collected from two digital pathology workflows. The evaluation of the multimodal architecture involves three tasks: WSI classification (on data from pathology workflow and from public repositories), multimodal data retrieval, and linking between textual and visual concepts. Noticeably, the latter two tasks are available by architectural design without further training, showing that the multimodal architecture that can be adopted as a backbone to solve peculiar tasks. The multimodal data representation outperforms the unimodal one on the classification of colon WSIs and allows to halve the data needed to reach accurate performance, reducing the computational power required and thus the carbon footprint. The combination of images and reports exploiting self-supervised algorithms allows to mine databases without needing new annotations provided by experts, extracting new information. In particular, the multimodal visual ontology, linking semantic concepts to images, may pave the way to advancements in medicine and biomedical analysis domains, not limited to histopathology.

Performance evaluation of multivariate deep-time convolution neural architectures for short-term electricity forecasting: Findings and failures

Deep learning (DL) models hold great promise in enhancing the decision-making abilities of electricity market participants and system operators in the short term, as they excel at capturing the intricate interdependencies and uncertainties in electricity consumption. Nevertheless, currently, the adoption of DL models for electricity prediction is not keeping up with the advancements in DL research. Within this framework, the aims of this study are to examine the probabilistic, multivariate, and multihorizon time series prediction of deep time convolution neural (DTCN) models. The effectiveness of DTCN in forecasting electricity consumption over daily, weekly and monthly time horizons is emperically assessed. We analyze both probabilistic and deterministic predictions. The evaluation data comprises real residential electricity consumption, average temperature and humidity with hourly granularity, collected over one year from a residential building. For evaluation, we considered accuracy and uncertainties, sensitivity to hyperparameters, the impact of exogeneous variables, as well as computational efficiency. Results show that the probabilistic models offer greater advantages when it comes to forecasting on multihorizons, especially for diverse datasets with non-uniform time series models. Unlike traditional models, the DTCN models demonstrated superior performance in capturing complex patterns and reducing forecast errors in short-term scenarios. However, their performance can be influenced by the sensitivity of hyperparameters. These findings can act as a benchmark for quantitatively assessing the performance of novel multivariate DL models in predicting electricity demand.