Proceedings of International Conference for ICT (ICICT) - Zambia

Engineering Education Management for Sustainable Development through Globalization, Diversity and Inclusion – A Curriculum Perspective

Bhagwan Das — 2023-12-03

In this paper, the conceptual framework is presented for Engineering Education Management based on Sustainable Development through Globalization, Diversity, and Inclusion. The framework will be helpful to revise and improve the existing curriculum for engineering students. Nowadays, globalization, diversity, and inclusion have had a significant impact on teaching and learning in the engineering curriculum. Their impact stimulates a broad perspective, cultural competence, and collaborative problem-solving skills that are vital for engineers working in an increasingly interconnected and diverse world. The paper discussed the various aspects of how Globalization, Diversity and Inclusion can be helpful to improve teaching and learning in the engineering curriculum. Our main contribution for this work is that we have proposed a methodology to bring Globalization, Diversity, and Inclusion in engineering education.

Machine Learning-Based Crypto Ransomware Detection Model On Windows Platforms

Martin Musonda — 2023-12-03

Ransomware, an evolving and highly destructive form of malware, presents substantial challenges in terms of detection and prevention. Despite extensive research and the application of Machine Learning (ML) models, existing defense mechanisms have struggled to provide complete protection, as most ML models fall short of achieving perfect detection rates. The study aimed to achieve several objectives related to Crypto- Ransomware detection. Firstly, it involved an examination of current ML frameworks employed in this field and the identification of associated challenges. Subsequently, the study focused on the creation of a new machine learning model designed for the detection and analysis of Crypto-Ransomware. By capitalizing on the shared behavioral patterns exhibited by ransomware, the proposed model attains an impressive 98% accuracy in recognizing ransomware on Windows systems. Lastly, the developed model's effectiveness in identifying Crypto-Ransomware was assessed through validation processes. Through evaluating multiple classifiers, the study identifies the Random Forest algorithm as the optimal choice for the model. This research marks a notable advancement in robust ransomware detection, working towards mitigating the far-reaching impacts of Crypto ransomware, a pervasive cyber threat.

Analysis of Breast Cancer Survivability Using Machine Learning Predictive Technique for Post-Surgical Patients

Tahsien Al-Quraishi — 2023-12-03

The primary objective of this study is to predict the likelihood of long-term survival for breast cancer patients who have received surgical treatment for a duration of five years or more. The aim is to provide healthcare providers with accurate predictions that can guide treatment plans and medication decisions. Despite existing breast cancer survivability prediction techniques, their accuracy remains low, limiting their practical utility. Additionally, there is a lack of research specifically addressing the survivability prediction of breast cancer post- surgery. Therefore, this study proposes a deep learning-based approach to predict survivability in this context. The effectiveness of the proposed model in predicting survival rates is evaluated using Haberman's survival dataset, obtained from the University of Chicago's Billings. Different evaluation measures, including accuracy, sensitivity, and specificity, are employed to assess the model's performance. Experimental results demonstrate that the proposed approach outperforms other models, achieving an accuracy of 83.18%, sensitivity of 85.54%, and specificity of 97.19%. The high accuracy of the proposed approach makes it suitable for use by healthcare professionals in predicting breast cancer survivability outcomes. It enables physicians to adjust treatments based on individual patient predictions. Consequently, the suggested method is advisable for practical implementation in systems designed to predict the survival chances of breast cancer patients after undergoing treatment.

Assessment of Biomass Energy Potentials and Appropriate Sites in Nigeria using GIS Computing Strategy

M.O. Ukoba — 2023-12-03

The geographical variability of biomass energy is an issue that requires the optima location of biomass energy facility. This paper presents a multicriteria GIS-based assessment of biomass energy potentials and appropriate siting of biomass plant in Nigeria. The study applies the weighted overlay multicriteria decision analysis method. Crop, and forest areas; settlement (energy supply areas); shrub/grass lands; barren land; waterbodies; distance from water sources; road accessibility; topography; and aspect are the criteria that were considered for locating the biomass facility in this study. The results suggest that the theoretical, technical and economical energy potentials of crop residues are highest in the north-east region of Nigeria and estimated at 1163.32, 399.73 and 110.56 PJ/yr, respectively, and least in the south-east at 52.36, 17.99 and 4.98 PJ/yr, respectively. The theoretical, technical and economical energy potentials of forest residues are highest in the north-west, estimated at 260.18,156.11 and 43.18 PJ/yr, respectively, and least in the south-east at 1.79, 1.08 and 0.30 PJ/yr, respectively. The GIS computing was able to identify the most suitable areas for siting biomass plants across Nigeria, in the Northern part of the country, and include; Niger, Zamfara, the Federal Capital Territory, Nassarawa, Kano, Kebbi, Kaduna and Borno state.

A Predictive Model to Support Decision Making for the Accreditation of Learning Programmes using Data Mining and Machine Learning

Francis Kawesha — 2023-12-03

Terabytes of data are produced by higher education institutions every year, and this data is crucial for determining how countries will develop. There are significant amounts of this Educational Data in a variety of relatively recent formats. We suggest a model for gathering, securing, and analyzing this substantial amount of data. The analysis of the data is used to evaluate the institution against a standard set by a Quality Assurance body, for the accreditation of higher education learning programmes. Therefore, the model supports the decision-making process in accreditation evaluation. The paper provides a proposed model using data mining and machine learning for the prediction of accreditation criteria, in the case of this paper the research considers academic staff appropriateness and adequacy.

An Analysis into factors affecting accuracy levels in deep learning models: A case of local language dataset in Zambia

Clement Mulenga Sinyangwe — 2023-12-03

Deep learning models are being trained to detect hate speech and abusive language using labeled examples. However, there are challenges, particularly in language dictionaries. Language dictionaries are collections of phrases and embeddings used to represent words as numerical vectors in a high-dimensional space. Collecting a high-quality dataset of words and their translations can be challenging, especially in low-resource languages with limited resources. Additionally, ambiguity and variation in language can make it difficult to accurately match words between languages. Out-of-vocabulary (OOV) words, which are not found in the training dataset and are unrecognized by the model, can also pose challenges when developing a local language dictionary, especially in low-resource languages with limited vocabulary. The main objective of this study was to analyse how the language dictionary affects the accuracy levels of deep learning models. CRISP-DM was used as a prefered mothodology. It was noted that in order for these challenges to be addressed, local datasets must be properly curated and preprocessed to guarantee that they are representative, diverse, and unbiased. The study was informed that cloud-based machine learning services can be used to overcome resource constraints and make model maintenance easier.

A Deep Hybrid Learning Model for Photovoltaic Solar Tracking Systems

Musa Phiri — 2023-12-03

The demand for clean and sustainable energy has increased the popularity of photovoltaic (PV) solar panels. PV solar panels are the most effective technology for transforming the sun's rays into a valuable energy source. To maximize the amount of energy captured, the rays of light from the sun must be at a 90-degree angle to the surface of the PV panel. To accomplish this, PV panels are modified with solar trackers that can effectively track the sun's position as it shifts during the day. Due to the randomness and nonlinearity of metrological data, using deep learning (DL) algorithms to enhance solar trackers has gained much popularity among researchers. However, given the varying nature of metrological data, single DL models sometimes fail to perform satisfactorily. For this purpose, this study applies sine and cosine transformations (SCT), a convolutional neural network (CNN), and long short-term memory (LSTM) to forecast the sun's trajectory. The SCT captures the cyclic components of the metrological data. The transformed data is then fed into a CNN-LSTM framework, where features of the sun’s movement are learned. The SCT-CNN-LSTM framework’s performance is evaluated concerning three other models based on the MAE, MAPE, and RMSE performance metrics. Based on the research findings, the CNN-LSTM framework has a superior performance, achieving an MAE, MAPE, and RMSE score of 0.0010, 73.5%, and 0.0012, respectively.

Exploratory Analysis and Preprocessing of Dataset for the Classification of Osteosarcoma Types

Amoakoh Gyasi-Agyei — 2023-12-03

Osteosarcoma is a born-forming tumor which is more common with children and young adults than adults. Classification of its type is crucial to its proper treatment and possible survival. Machine learning models, trained on datasets of the disease, are are effective classification tool than hand-crafted features which are highly dependent on a pathologist’s expertise. However, machine learning models are only useful if the dataset used to train them are representative, of good quality and well prepared. Thus, data preprocessing and statistical analysis of datasets used to train models are necessary precursors to model learning. Data preprocessing is the most demanding task in the model learning pipeline. Thus, availability of a pre-processed quality dataset for a given task is desirable for model learning tasks. Two things are needed to obtain good results in a machine learning project: good data preprocessing and good algorithms. This paper provides a thorough preprocessing and statistical analysis of a 1144-sample dataset of osteosarcoma patients, to render the dataset ready for model learning. The efficacy of the preprocessing methods is verified by training multiclass logistic regression in Python using datasets with 63 of the 69 variables, with PCA and feature selection to achieve the respective predictive accuracies of 19.27%, 65.14% and 80.28%.

OCCUPATIONAL HEALTH & SAFETY INFORMATION MANAGEMENT SYSTEM USING DISTRICT HEALTH INFORMATION SOFTWARE

Timothy M Lwiindi — 2023-12-03

Generally, there are insufficient studies on occupational health and safety in Africa. This study examines how District Health Information Software (DHIS) can be used as a health information systems infrastructure tool for occupational health and safety. Health Information System data is vital, yet most developing countries' health data collection, collation, compilation, analysis, and reporting is inadequate, inaccurate, and untimely, making it unusable for decision-making. The goal is to examine how various user perceptions affect work-related security and health behavior. The objective is to conduct a baseline study to examine New Partnership for Africa's Development (NEPAD’s) challenges in sharing occupational health and safety data with regional partners. This report investigates DHIS as an information infrastructure. The study also identifies factors that affect Occupational Health Information System Management acceptability and utilization using the Unified Theory of Acceptability and utilization of Technology (UTAUT) Model. The researcher used a descriptive research approach in which questionnaires were distributed to a randomly selected group. The acquired data was analyzed using descriptive and correlation analysis in the social package for statistical sciences software (SPSS). It used the descriptive analysis, to find a relationship between individuals’ intention to use Occupational Health Information Management System and their belief that the system would improve their performance and efficiency in carrying out their tasks. The study will conclude the hypotheses stated, Effort Expectancy, Performance Expectancy and Social Influence significantly influence behavioral Intentions for users to use occupation health information management system.

Visualization of the effects of forest bathing: Comparison between Japan and Thailand

Atsushi Ito — 2023-12-03

Recently, health tourism has become popular worldwide. One category of health tourism is Shinrinyoku (Forest Bathing). Shinrinyoku was born in Japan in 1984. There are several benefits of Shinrinyoku, such as relaxing and refreshing. We have been studying the source of the effect of Shinrinyoku in Oku-Nikko. In February 2023, we researched Shinrinyoku in Kaoyai National Park in Thailand. In Oku-Nikko, we found that we can be relaxed near a waterfall or river. In the forest of Thailand, we found that not only near a river and waterfall, walking in a rain forest was also effective for relaxing.

Using ArcGIS for Optimal Location of Small Hydropower Sites in Nigeria for Sustainable Development

Benneth Oyinna — 2023-12-03

This paper focuses on the prospects of small hydropower plants (SHP) in Nigeria and utilizes ArcGIS software for analysing the country's hydropower energy potential. The analysis conducted using ArcGIS reveals the significant SHP potential in various states across Nigeria. By overlaying water areas and waterline data on maps, potential sites for SHP are identified, particularly in states such as Borno, Niger, Edo, Anambra, and Jigawa. Further analysis was done using data for water lines in Nigeria converted into shaped files for the six geopolitical zones of Nigeria, with the various states and local government areas, to provide expanded views for different possible schemes for SHP. Data set were built up for the different geopolitical zones and statistical analyses were done for SHP potentials. Inverse Distance Weighting (IDW) tool on ArcGIS was used for interpolations to indicate places that are good for locating dam schemes for hydropower production as well as other schemes that require less water storage for small hydropower production. Additionally, it was determined which zones in which each state excelled in terms of availability of inland waters and lands susceptible to inundation. There were also some significant difficulties and potential in evaluating Nigeria's small hydropower projects using ArcGIS.

Modelling and Analysis of Transient Evoked Otoacoustic Emissions for Human Biometric Recognition

Johnson I Agbinya — 2023-12-03

This paper reports on application of data analytics techniques for determining the status of human ear, variation in ear conditions and the specifics of whose ear they are. Principal component analysis of transient evoked otoacoustic emissions from human ears were derived based on recording of responses to external audio excitations undertaken fifteen months apart. Results indicate the method established in the paper is suitable for person recognition and for identifying when deterioration of the hearing performance of an ear has taken place. Ear transformation matrix is introduced. The transformation matrix represents a scaling of the eigenvectors using a Hermitian matrix or scaling matrix. They are however known to be the eigenvalues obtained from the PCA analysis. While the eigenvalues could be seen to represent audio loudness scaling, the eigenvectors represent further ear deterioration. Eigenvalues are maintained when the audio performance of ear to external excitation has not changed. Variation of the Hermitian is also variation of the ear condition. The lengths of the eigenvectors are considered as estimates of the change in ear loudness performance as it can be seen as equivalent to the power content of the eigenvectors.

Automatic Classification of research grants proposals using a multi– class machine learning model

Rebecca Lupyani — 2023-12-03

Research and Development has become fundamental to the economic development of every nation. Many countries have established institutions to promote, support and fund research and innovation in societies. These institutions seek to monitor and keep track of how much money they are willing to or have been spending on research and development activities in specific fields or topics. Funding investment decisions are based on whether proposed research ideas fall under disciplines of interest to national development. It is therefore imperative that research proposal documents submitted for funding consideration are classified according to respective disciplines. This paper explores the adaptation of the text classifier Support Vector Machine (SVM) for multi-classification and use it to automatically classify scholarly research documents and predict eligibility of funding. The experiment results demonstrate that the SVM model performed well with an accuracy performance of 89%. The study recommends implementing Application Programming Interface (API) endpoints for the model, to facilitate its integration with third-party tools and services to automatically classify the research proposal documents and award research grants.

SENSING WATER POLLUTION IN THE KAFUE RIVER USING CLOUD COMPUTING AND MACHINE LEARNING

Mumbi Mumbi — 2023-12-03

Clean water and sanitation are the sixth goal under the UN Sustainable Development Goals. However, reports have shown that over 129 countries are not on track to reach this goal by 2030. Besides the lack of basin management, countries are behind on monitoring of the water bodies. Water pollution affects the livelihood of people living in the catchment area of the river especially the people living in the rural areas and the animals that are dependent on that water. People who live in rural areas do not have the privilege of a piped water network that has treated water. Currently, in Zambia, water is monitored once every quarter and so, this leaves the water unmonitored for most of the time. This research proposed the development of a model based on IoT, Cloud Computing and AI for data collection and monitoring, and developed a prototype based on this model which uses machine learning to predict the quality of water. A water monitoring device was built using sensors, an Arduino and a Raspberry pi. The sensors used measured pH, temperature, electrical conductivity, total dissolved solids and turbidity. An Artificial Neural Network with one hidden layer was used to predict the Water Quality Index. This index was based off the National Sanitation Foundation Water Quality Index (NSF-WQI). The results of the model showed that it had an R2 score of 0.953, MAE and MSE of 0.835 and 1.280 respectively. These results support the use of an ANN in the predicting WQI

Feature Enhancement and Chaining of Deep Nueral Networks in Colorectal Cancer Classifaction based on Gut-Microbiome Data

Mwenge Mulenga — 2023-12-03

Colorectal Cancer (CRC) is among the top three cancers in world. The current clincal methods for CRC detection have several limitations which range from low accuracy, discomfort and high costs. Availability of next generation sequencing (NGS) technology has opend an opportunity for non invasive detection of CRC which uses gut-microbiome abundance in stool samples. The high dimension of sequence base microbiome data has prompted research interest in the application of machine learning (ML) in order to classify host disease based on microbial counts. However the classification performance of ML methods such data is still limited by factors shuch as high dimensionality and data imbalance. Therefore, in this paper, we propose a deep nueral network based method that combines feature extension and feature and chained execution of deep neural network to improve CRC classifaction based on gut microbiome in stool samples. The proposed method scored a mean area under the receiver operating characteristics curve (AUC) of approximately 95.4%, which is higher than state-of-the-art methods. The proposed method can positively contribute to the development of robust diagnostic and prognostic methods for CRC.

Large-Scale Analysis of Medical Image Metadata

Elijah Chileshe — 2023-12-03

This paper offers an ongoing exploration of the systematic analysis of Medical Image Metadata encoded using the Digital Imaging and Communications in Medicine (DICOM) standard. The paper carefully looks at the organization of diverse medical image data, the complex navigation of ethical considerations involving data authorization and anonymization and the subsequent intricate processing and metadata extraction procedures. The research is part of a larger project that aims to address the pressing need for radiologists in the Republic of Zambia. With only a limited number of trained radiologists available to serve a significant population, innovative solutions are urgently required. Simultaneously, this study explores the potential of streamlined medical imaging workflows through the application of enterprise imaging techniques. The paper is dedicated to providing detailed insights into the methodologies that support the execution of large-scale medical image metadata analysis. By capturing the collection of images from multiple sources, addressing ethical concerns for patient privacy and extracting metadata from DICOM files, this ongoing study continues to provide valuable insights into the refinement of medical imaging practices and the enhancement of clinical decision-making processes.

he Readiness, Risks and Mitigation Measures of Quantum Supremacy to Current Digital Security Measures and Infrastructure

Mulima Chibuye — 2023-12-03

With the growing funding and research into quantum computers by countries such as the United States, China and India, there is growing concern about the security of modern encryption systems that the digital space relies on. It has been shown that Shor’s algorithm is capable of breaking Rivest–Shamir–Adleman encryption which forms a fundamental part of online security today. This paper explores the readiness of countries and organizations to cope with Quantum Supremacy which is expected to be achieved by the year 2033 and potentially sooner as the estimates given are based on the current technology but as can be noted, it is very hard to predict breakthroughs. However, scientists today are quite aware that fully functional quantum computers are not far off. Through a literature review using resources such as Google Scholar, IEEE Xplore, and the ACM Digital Library, we will investigate the current research and progress into the field of Quantum Computers, the initiatives to make global infrastructure ready for the expected breakthroughs and the implications, risks, and mitigation measures currently possible to prepare for Quantum Supremacy. We also look at the Technology Acceptance Model as relates to the adoption of Quantum Computers. The study will focus on the measures that countries and organizations have put in place to protect digital space and digital data. Within the scope of the analysis, we will apply the Pareto principle on the top 20% of the wealthiest Nations and Organizations and on 20% of the wealthiest middle-income nations to provide a broad overview of readiness. We will also provide a case study of the United States which is known to be a technological Superpower. We will analyze efforts to prevent unlawful access by both private entities and sovereign nations. This work aims to contribute to the ongoing assessment of global readiness for quantum supremacy.

Diversity Measure to Tackle the Multiclass Problem in IoT Intrusion Detection Systems

Osama A. Mahdi — 2023-12-03

The advent of the Internet of Things (IoT) has instigated transformations in various domains, such as healthcare, smart homes, agriculture, transportation, and manufacturing. With the swift proliferation of IoT networks, new security challenges have surfaced, exposing them to a plethora of attacks. To counter these, machine learning-driven intrusion detection strategies have been introduced, which scrutinize the behavior and communication patterns of IoT devices to identify and nullify any suspicious undertakings. While these methodologies demonstrate high accuracy and minimal false alarm rates in static scenarios, their performance stability in dynamic, evolving environments remains undetermined. One critical issue pertains to multiclass problems, wherein the complexity of diverse attack types can significantly affect the efficacy of machine learning-based intrusion detection systems, if not promptly recognized and addressed. This paper introduces an innovative IoT Intrusion Detection approach that incorporates the Diversity measure as a model drift detection method to tackle the multiclass problem in IoT networks. Our proposed approach can detect previously unknown attacks in IoT networks through an advanced drift detection technique.

A Facial Authentication-based Deepfake Detection Machine Learning Model

Joseph Mwanza — 2023-12-03

In an era dominated by digital media, the escalating menace of media distortion, particularly propelled by the advancement of deepfake technology, has emerged as a critical concern spanning the realms of virtual landscapes and reality. The rise of deepfake technology has posed significant challenges to the authenticity of visual content in today's digital world. This study proposes a novel approach to deepfake detection using pixel analysis. By closely examining the pixel characteristics and patterns within manipulated images and videos, we developed an algorithm that can distinguish between real and fake content with high accuracy. Our algorithm combines two state-of-the-art deep learning models, Resnext and Long-Short Term Memory (LSTM), in a supervised machine learning framework. To enhance the performance of our algorithm, we applied standard pixel normalization during the preprocessing phase. Our proposed method achieved an impressive accuracy score of 95.6% on a public dataset of deepfake images and videos. This result demonstrates the efficacy of pixel analysis in detecting deepfakes. This research contributes significantly to countering the increasing threat of deepfake media manipulation, safeguarding the authenticity of visual content in today's digital world.

A Review on Machine Learning Applications in Localization in 5G and Beyond Wireless Communications

Agburu O. Adikpe — 2023-12-03

Location is the process of estimating the position of a device or user in a wireless network. Location is critical for many applications and services in 5G and beyond wireless communications. However, localization faces many challenges in complex and dynamic environments, such as multipath propagation, non-line-of-sight (NLOS) conditions, and limited bandwidth and power resources. Machine learning (ML) is a promising technique that can improve location performance and efficiency by exploiting the large amount of data available in wireless networks. In this article, state-of-the-art ML localization techniques are reviewed. In addition, recent ML localization techniques are compared, and observations from the comparison are delineated. Additionally, challenges with possible recommendations are presented.

Towards Leveraging AI Deep Learning Technology as a means to Smart Farming In Developing Countries: A case of Zambia

Barbara Chalwe Kunda Kunda — 2023-12-03

Despite Plants being a major source of food for the world population, they continue being ravaged by plant diseases, a situation that greatly contributes to significant decline in production, which ultimately adversely impacts on food Security. Owing to the fact that manual plant disease monitoring is both laborious and error-prone, there has been a heightened need for farming practices that are sustainable, efficient, reliable and cost effective, necessitated by the need to adopt cutting edge technologies such as Artificial intelligence. Smart farming as an innovative approach to agriculture, offers farmers in developing countries a means to effectively diagnose and proactively manage plant diseases. Innovative smart farming solutions through the use of technology makes precision in agriculture possible, by enabling farmers to adopt practices that are optimized based on real-time data and analytics. The increased precision in dealing with problems such as crop disease detection help reduce consumables during disease maintenance, thereby increasing profitability and enhancing food security. This study aims to leverage technology as a means to smart farming in Zambia, a developing country in the sub-Saharan region of Africa, by employing a Convolutional Neural Network (CNN) model for the detection of tomato leaf diseases. Tomato production in Zambia faces significant challenges due to the prevalence of diseases such as early blight, late blight, and leaf mold, which can potentially lead to substantial crop losses. Nonetheless, Early and accurate detection of these diseases is crucial for effective management and increased productivity. This study proposes the use of Automated tomato leaf disease detection through the use of a convolutional neural network model. The plant village dataset which is one of the largest open access repository of expertly curated leaf images for disease diagnosis is used in this study. The CNN model is trained using this dataset, enabling it to learn discriminative features and patterns associated with different disease classes. This system offers an opportunity to empower farmers with timely and accurate information regarding disease occurrence and severity, enabling them to take proactive measures for disease management. By leveraging technology as a means to smart farming, the study aims to improve the efficiency, productivity, and sustainability of tomato farming in Zambia. The Convolutional neural network for the detection of tomato leaf disease was built, successfully trained and deployed. The accuracy of the CNN Model was at 95.8%

COVID-19 CONTACT TRACING USING ACCESS CONTROL AND FACEMASK RECOGNITION

Azwel Simwinga — 2023-12-03

Since the outbreak of the Covid-19 pandemic, Facial Recognition technologies have experienced rapid and extensive adoption worldwide. Initially designed for regulating access to specific facilities and ensuring compliance with mask-wearing protocols, these systems now require enhancements to extend their utility beyond the pandemic. This study aims to augment an existing facemask detection system by incorporating future-proof functionalities, particularly Facial Recognition. The inclusion of access control features, such as Facial Recognition, seeks to advance the system's capabilities, allowing for improved user identification and access management. A TensorFlow Lite machine learning facemask detection model was developed, utilizing a dataset collected from GitHub and Kaggle. The dataset consisted of 5,092 photos categorized into three groups: "with_mask," "without_mask," and "mask_worn_incorrect." To ensure accurate model performance, 70% of these images were allocated to the training set, while the remaining 30% were assigned to the test set. Subsequently, a Python application was created to incorporate this robust facemask recognition model. Notably, the Python application goes beyond mere facemask detection by incorporating Facial Recognition capabilities. The Facial Recognition functionality was implemented using the haarcascade_frontalface_default algorithm. Deployed on a Raspberry Pi 4 edge device, the Python application streamlines user registration by assigning each participant a unique ID based on their National Registration Number (N.R.C). The integration of Facial Recognition technology strengthens the system's ability to accurately identify individuals, reducing the chances of impersonation or unauthorized access, while also enforcing facemask regulations. The proposed solution contributes to the ongoing efforts to create safer and more efficient access control systems in a post-pandemic world.

Comparative Analysis of IoT Architectures for Mine Environmental Monitoring: A Literature Review and Evaluation

Sipiwe Chihana — 2023-12-03

The viability of mining operations and the safety of communities depend on environmental monitoring. This study looks at different Internet of Things architectures that are used to monitor the environment in mine environments, focusing on factors such as communication dependability, energy utilization, scalability, real-time observation, and cost effectiveness. The paper identifies problems with current monitoring methodologies, such as inconsistent monitoring and expensive maintenance costs, based on an examination of some scholarly articles and a case study with the Zambia Environmental Management Agency (ZEMA). As a response, the research proposes a hybrid IoT framework that combines cloud services, remote calibration features, and remote sensing networks while also factoring in climate adaptation. These findings highlight the revolutionary potential of IoT systems in boosting environmental surveillance in mine environments, which can result in sustainable practices and improved community health. These insights are vital for regulators and other interested parties.

Software Tools for Supporting Automatic Interpretation of Medical Images

Andrew Shawa — 2023-12-03

In the domain of medical imaging, the role of automated image interpretation tools is becoming increasingly critical in facilitating the diagnosis and treatment of diverse diseases. The escalating volume and intricacy of medical images necessitate the development of advanced tools that can support automatic image analysis. This paper outlines on-going work associated with the design and implementation of extensions and plugins for widely used DICOM viewers, specifically Weasis, Dicoogle, and Orthanc. The primary objective is to augment the functionality of these viewers, empowering them to assist radiologists and healthcare professionals in the comprehensive interpretation and analysis of medical images. This abstract outlines how DICOM viewer extensions and plugins can be integrated with machine learning models to enhance the efficiency and accuracy of medical image interpretation, ultimately leading to improved patient care and outcomes.