Browse Items (3095 total)
Sort by:
-
An exhaustive examination of automatic speech recognition
Speech is nature gift to the human being to differentiate with other creatures on the earth. Speech research stands out as one of the most daunting fields amidst numerous challenging research domains. This current analysis of existing literature aims to provide insights for future endeavors within the global speech research community. The study delves into various challenges associated with speech corpora, front-end algorithms aimed at efficient speech representation, and back- end engines tasked with the recognition process. Thirteen speech corpora undergo scrutiny concerning factors such as language diversity, duration, developmental progress, and accessibility. Furthermore, this research review illuminates potent methodologies that foster the extraction of rich features and bolster robust speech recognition capabilities. This gives an idea on how various methods are available to recognize speech in an effective way. 2026 Author(s). -
IoT based hybrid patient health monitoring system
According to the data released by the World Health Organization (WHO). Cardiovascular diseases (CVDs) are still the leading thread worldwide. Every year around 17.9 million fatalities worldwide are attributed to CVDs, making up 31% of overall deceases. Heart attacks and strokes are the main causes account for most of these deaths, and a large fraction happen so early in the age group of under 70. Check-ups are essential to monitor the healthcondition of the elderly people, which generates a considerable challenge to the existing medical field. Subsequently,It's becoming more crucial to diagnose diseases quickly and accurately with an affordable cost. The World's population growth need for a smart and affordable healthcare solutions to reduce the medical costs. Thus, the development of an effective health monitoring system that can quickly identify irregularities in health and provide precise diagnoses based on gathered data is imperative. New developments in cloud computing and mobile technologies have led to the development of a number of cloud-based healthcare products and services. These cloud systems allow for the automatic collection and transmission of medical data to providers from any place, allowing for the network-based delivery of patient feedback. This project's goal is to use the ThingSpeak IoT platform to create and implement an Internet of Things-based patient health monitoring system in order to meet these goals. 2026 Author(s). -
Advancements and challenges in deep learning for breast cancer screening: A review
Breast cancer continues to be the prevalent cancer on a global scale, playing a major role in the worldwide cancer statistics, the critical role of early detection in reducing death rates is underscored. In the context of breast cancer, screening, deep learning (DL) emerged as a game-changer, providing notable improvements over existing techniques. This review explores the use of DL in analysing images from various sources such as X-rays, ultrasound, magnetic resonance imaging, and biopsies. Additionally, it highlights DL's potential to pre-screen for cancer by integrating diverse data, including demographic information, biological markers, and meta-analytical risk assessments. The analysis reveals that deep learning frameworks, especially those optimized with feature selection techniques, attain the minimal false-negative rates, effectively distinguishing between patients with and without cancer. Notably, DL models demonstrate lower prediction uncertainty compared to traditional machine learning, as shown by reduced standard deviations in performance metrics. Additionally, the paper proposes a cascade network model that achieves 98.61% classification accuracy and a 98.41% F1 score by segmenting tumours with a UNet architecture and classifying them with a ResNet backbone. Despite these advancements, challenges such as limited annotated data and adaptability to new data domains persist. In response to these issues, the proposed Self AdaptNet leverages innovative self-supervised learning and adversarial techniques to improve the resilience as well as adaptability of BC detection models.AI technology, particularly DL-based systems, has the capacity to completely transform breast cancer screening by improving screening accuracy and reducing observer variability. However, clinical adoption requires standardized guidelines, trustworthy AI practices, and collaboration among researchers, clinicians, and regulatory bodies. 2026 Author(s). -
High-precision lung disease detection and classification from chest radiographs using deep and ensemble neural networks
Chest X-rays are a quick and effective way to diagnose lung diseases. This research developed deep learning models to automatically detect chest X-rays of COVID-19, normal, and viral pneumonia patients. The goal was to evaluate deep learning for automated detection of lung diseases from chest X-rays. The research implemented transfer learning with ResNet101 and EfficientNetB0 architectures using a public chest x-ray database with over 21,000 images across COVID-19, normal, and other pneumonia infection classes. Pretrained ImageNet weights were used to initialize the models before fine-tuning them to classify features in chest X-rays. Data augmentation techniques like rotation, shifting, and flipping were applied to expand the number and diversity of training images. The models achieved exceptional performance with accuracy scores of 93.7% for ResNet101 and 95.3% for EfficientNetB0 on test data. Additionally, an Ensemble model, the combination of the two models, was implemented, achieving an accuracy of 96.4%. The findings demonstrate the capability of Ensemble deep convolutional neural networks for accurate automated classification of chest X-rays for Lung disease. Through data augmentation and transfer learning, high-precision models were developed without needing exceedingly sizeable medical image datasets. These deep learning classifiers could serve as rapid diagnostic decision support systems to identify potential lung disease patients using readily available chest X-rays. Such tools could assist healthcare providers, especially when access to expensive diagnostic tests is limited. 2026 Author(s). -
Deep learning-based diabetic retinopathy detection with advanced image segmentation and transfer learning techniques
Diabetic retinopathy (DR), a dangerous side effect of diabetes, can result in permanent blindness. This work presents a state-of-the-art deep learning-based system that uses retinal images to detect and classify DR early on. Utilizing transfer learning and pre-trained models, the system combines Django, Numpy, and Keras to improve diagnostic precision. It accurately detects DR-affected areas and delivers real-time graphical outputs for prompt medical interpretation and decision-making using the ResNet and Mask RCNN architectures. Simple picture uploads are made possible by the user-friendly interface, which lets Numpy handle data processing and preparation. To improve accuracy and reduce the amount of new data required, the system uses transfer learning and pre-trained datasets. The system's robustness and efficacy are highlighted by its evaluation, which shows its high accuracy with an overall accuracy of 95.55%, precision, recall, and F1-scores above 0.95. The suggested approach provides an affordable, effective, and scalable means of detecting DR early on; it is especially helpful in healthcare settings with limited resources. The technology has the potential to greatly enhance patient outcomes and lessen the toll that diabetic retinopathy has on both individuals and healthcare systems by enabling prompt diagnosis and treatment. 2026 Author(s). -
Analysis of Biometric Systems for Secure Human Recognition
In the realm of contemporary computing, the recognition of humans has emerged as a crucial element, finding utility in both mundane daily tasks and sophisticated operations across diverse IT applications. The process of identifying individuals often involves harnessing their distinctive biological, chemical, and behavioral attributes. To achieve this, a biometric system, functioning as a computer-based automated mechanism, is employed to authenticate and confirm alter the perspective or framing of users by exploiting their biological characteristics. In present-day applications, an existing entity exists notable emphasis on the biological aspects of individuals as the primary means of identification. While utilizing the chemical traits of humans for identification yields greater accuracy and reliability, practical implementation proves to be challenging. This article presents the execution or outcome of automatic human recognition systems derived from diverse sources or perspectives parameters such as user psychology, ease of use, security, reliability, and market share. The results suggest that these systems offer authentication and recognition capabilities, but it is noteworthy that the security of these systems at the template level poses a significant challenge for designers. 2025 Author(s) -
Chain funds: Transforming healthcare crowdfunding through blockchain technologies
Crowdfunding in healthcare refers to the practice of raising funds from a large number of individuals through online platform. This will eventually support the various medical expenses involving hospitalization and treatment. This method enables individuals with health problems or an urgent need for medical treatment to go out and seek financial assistance from a large number of people. Crowdfunding on health is now widely recognized as the means of catering for medical bills, surgeries, experimental treatments and other health-related expenses. The research presents a decentralized blockchain based crowdfunding platform built using ReactJS, Solidity and Thirdweb SDK. This new platform aims to change traditional crowdfunding techniques through utilization of the benefits of blockchain such as transparency, security and automation provided by smart contracts. The user interface has been designed using ReactJS; creating smart contracts on the Ethereum Blockchain was done under Solidity; Thirdweb SDK was used to create a link between the system and the blockchain. Some significant features of this platform involve start-up campaigns organization, contribution management, milestone tracking and automatic distribution of funds in order to enhance customer satisfaction and minimize operational costs. Moreover, users are awarded with certificates and crypto tokens. In addition, this policy ensures high level security that comes with immutability in blockchain technology by eliminating middlemen hence live monitoring of fund is possible through it at all times and disbursements. Rigorous testing has been conducted to assess performance, security, and scalability, highlighting the advantages of decentralization in contrast to centralized crowdfunding models. 2025 Author(s). -
Thermal performance evaluation of vertical slotted circular fins over turbulence flow regime in a heat sink
The study investigates the thermal performance of vertical slotted circular fins in heat sinks under turbulent flow conditions. As electronic devices become more compact and power-dense, efficient thermal management is critical. This research uses numerical simulations to evaluate six different slot sizes (0.5 mm to 3 mm) in terms of their impact on heat transfer and flow dynamics. The fins, modeled in 3D and subjected to Reynolds numbers ranging from 8490 to 23300, were analyzed for heat transfer efficiency and friction factors. Results indicate that slotted fins outperform solid fins, with the S-D slot configuration achieving a 17.73% increase in the Nusselt number and a 28.77% reduction in friction factor at a Reynolds number of 13,365. The Thermal Evaluation Criteria highlight the S-D slot as the most effective, providing a 15.17% improvement in overall performance. These findings underscore the potential of slotted fins in optimizing thermal management systems by balancing enhanced heat dissipation with reduced energy consumption. 2025 Author(s). -
Skin cancer prediction using AI: A bibliometric analysis
Skin cancer is a major public health concern globally, with early detection being crucial for successful treatment and management. Artificial intelligence (AI) has emerged as a promising tool for aiding in the early detection of skin cancer [15, 19, 23, 41]. This paper conducts a literature review and bibliometric analysis to explore the current landscape of AI-based skin cancer prediction. This bibliometric analysis systematically examines the landscape of research on skin cancer prediction using AI. The aim of the study is to identify the research trends, keyword contributors, influential authors, and research hotspots [13, 31]. Through this bibliometric analysis, this study offers insights into the evolution of AI-based approaches for skin cancer prediction. By producing and analyzing bibliometric data from relevant scholarly publications, this study provides a comprehensive overview of the current state of research in this domain, informing future directions for advancing skin cancer prediction using AI technologies. 2025 Author(s). -
A bibliometric analysis of fruit disease prediction using machine learning
In recent years, there has been a growing interest in leveraging machine learning techniques for the early detection and prediction of diseases affecting fruit crops. This study presents a comprehensive bibliometric analysis of research literature focused on fruit disease prediction using machine learning algorithms. Through systematic review and analysis of a large corpus of scholarly articles, conference papers, and patents, this paper aims to provide insights into the current trends, key research themes, influential authors, and popular machine learning methods in this domain. This paper conducts a literature review and bibliometric analysis to explore a significant increase in research activity in fruit disease prediction using machine learning, indicating the increasing importance of this area in agriculture and plant pathology. Various machine learning and deep learning algorithms, including convolutional neural network (CNN), decision trees, random forests and LSTM have been widely employed for disease prediction tasks. Moreover, the study identifies common datasets, evaluation metrics, and challenges encountered in this field. Overall, this bibliometric analysis provides valuable insights for researchers, practitioners, and policymakers interested in fruit disease prediction, highlighting opportunities for collaboration, innovation, and advancement in agricultural technology and plant health management. 2025 Author(s). -
Comparative analysis of machine learning algorithms for predicting student success and enhancing their educational outcomes
The primary objective of this study is to predict the performance of students and evaluate the efficacy of various machine learning algorithms in predicting student success based on their marks and grades (academic factors). Through a comprehensive review of literature and experimentation, this research compares the performance of different machine learning models, including but not limited to decision trees, random forests, support vector machines, logistic regression, and neural networks. The evaluation metrics considered in this comparative analysis include accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC-ROC). Fourteen experiments have been performed and preliminary results suggest that performances of students on the basis of academic factors might be predictable and by understanding the strengths and weaknesses of student's educational outcomes and foster student achievement can be improved. Through extensive experimentation and comparative analysis, XGBoost(ExtremeGradient Boosting) and AdaBoost demonstrated as the most effective predictive models to analyze the students' performance. 2025 Author(s). -
Survey on deep learning techniques used for object identification of underwater forward looking sonar images
Underwater object identification using forward-looking sonar (FLS) images is crucial for autonomous underwater vehicles (AUVs) for navigation and obstacle avoidance. Deep learning techniques have emerged as powerful tools for object recognition in various domains. This paper surveys deep learning approaches employed for object identification in FLS images. We examine the effectiveness of popular deep learning frameworks such as YOLOv5, EfficientDet, and MobileNet, and transfer learning, data enhancements to improve object recognition performance, and the role of adversaries training. We also examine the potential of focusing and lightweight CNN algorithms developed for FLS images despite these advances, challenges still exist due to the limited number of registered cases. The paper analyzes how deep learning methods address these challenges and highlights their effectiveness in object identification. We aim to provide a comprehensive overview of the current state-of-the-art in deep learning for FLS object identification, paving the way for further research and development in this field. Results of this study show that the proposed algorithms improve obstacle detection accuracy and processing speed of sonar images. At the same time, the proposed algorithms ensure AUV navigation safety in a complex obstacle environment. 2025 Author(s). -
Novel mammography images approach for breast cancer diagnosis using ensemble feature extraction
By using ensemble feature extraction methods to mammography pictures, this study introduces a novel strategy for the early detection of breast cancer. Beginning with preprocessing stages that use data augmentation to improve the dataset, the technique incorporates a methodical flowchart. Following the creation and individual training of an ensemble model that incorporates CNN architectures like as DenseNet, AlexNet, and i-Alex, the final model attains an impressive level of accuracy. Optimized feature vectors are the end result of a process that begins with feature fusion and continues with dimensionality reduction methods like principal component analysis (PCA). Utilizing LASSO and ReliefF for feature selection helps to refine the collection of features, which in turn improves accuracy metrics. Utilizing cross-validated hyperparameter optimization, classifier training showcases the effectiveness of SVM, Random Forest, and XGBoost. The ensemble method is clearly better according to the performance assessment, which takes into account sensitivity, specificity, F1-score, and AUC. Integrating the chosen classifier into a mammography screening system ensures clinical interpretability by providing clear visualizations. Updating the model with fresh data on a regular basis and doing continuous monitoring ensure that it remains accurate. By working together in the clinic and taking radiologists' comments into account, we can improve the system's performance and reveal its capabilities as a cutting-edge instrument for accurate breast cancer detection. 2025 Author(s). -
Prediction of CO and NOx Emission from Gas Turbine Using Machine Learning
In gas-turbine-based power plants, predictive emission monitoring systems (PEMS) are used to validate and back up the expensive continuous emission monitoring systems. Increasing energy consumption increased deforestation and carbon and flue gas emissions, harming the environment. The availability of relevant and ecologically sound data is crucial to their successful deployment. In this article, we adopted the Gas Turbine CO and NOx Emission Data Set Data Set from UCI machine learning repository to predict the CO And NOx emission from gas turbine using machine learning (ML). We developed the model using random forest and support vector algorithms. The random forest algorithm performs better for the data. 2025 Author(s). -
Enhancing Video Surveillance for Crime Detection Using Anomaly Detection Techniques
Security cameras are widely used to detect and prevent crimes, but the number of surveillance videos has increased due to this prevalence. The process of detecting similarities or data points that significantly depart from the norm or expected behavior of a given system is known as anomaly detection. Predictive maintenance, network intrusion detection, and fraud detection are just a few of the areas where anomaly detection is applied. By processing these videos with the help of a suitable machine learning algorithm, unfavorable events can be brought to the attention of experts to manually monitor. Since these unfavorable events are of various types and few in number, this problem can be addressed in the anomaly detection structure. An anomaly detection algorithm has been developed using the UCF-Crime dataset consisting of 1900 surveillance videos of various lengths. In this context, video surveillance refers to observing the scenes of improper human behaviors which are termed as real world anomalies. Depending on the availability of data sets, anomaly detection algorithms can be supervised, unsupervised, or semi- supervised. The quality of the data and the selection of the best algorithms determine how well anomaly detection techniques work. This paper proposes the use of anomaly detection techniques to enhance video surveillance systems for crime detection. By identifying unusual activities in surveillance footage, the system can alert authorities to potential criminal activity and improve overall security measures. The effectiveness of this approach is demonstrated through experiments and analysis of real-world surveillance data. 2025 Author(s). -
Customer churn behaviour prediction in telecommunication using classification algorithms and modelling
The cost of obtaining a high-quality client is usually five times more than the cost of keeping an existing customer. This is why it is very important that businesses keep their customers at home. To retain and improve their customers' satisfaction, researchers in various fields such as marketing, information technology, and business intelligence studied various ways to deliver the best possible services. Despite the good performance of the work done before, there is still a considerable gap in their prediction of the churners. In most cases, the training dataset is too large, and the high dimensionality of it causes the classification algorithms to fail. In the present paper, an attempt was made to estimate customer churn with greater accuracy in the membership of cellular wireless services using a call details records dataset consisting of 3333 clients having 21 attributes each. With the advancement of Machine Learning (ML) and artificial intelligence, most popular approaches such as logistic regression, CART, and C5 algorithms have been used with and without using the data balancing technique SMOTE. The performance evaluation of these predictive models is done using the model accuracy, confusion matrix, AUC value, ROC curve, and Cohen's Kappa statistics. The study results indicate that the C5 algorithm could estimate customer churn with an accuracy of more than 92% for both balanced and imbalanced datasets. 2025 Author(s). -
A study of failure prediction of Indian banks using various machine learning algorithms - An examination of predictive accuracy
Banks play a key role in strengthening the economy; hence their survival is very important. It is necessary to evaluate the failure probability of banks correctly based on the factors associated with it. Over half of the assets in the financial sector in India are held by the banking sector, which holds a strong position. Phased implementation of financial sector reforms has resulted in an exciting moment of rapid transformation for Indian banks. The study here focuses on the establishment of machine learning approach to compute and compare the extent of bankruptcy based on the accuracy measure-Support Vector Machine classification, Random Forest, Logistic Regression, Nae Bayes classification using the data of 250 Indian banks having qualitative variables from 2015 to 2020. The feature selection in this paper is based on correlation and relief algorithms. The explanatory features of the dataset are drawn by implementing a two-step feature selection technique and the selected features are fed and further used for prediction using the Random Forest technique, Logistic Regression, Support Vector Machine, and Nae Bayes classification techniques. The results reveal, that the support vector machine shows a score of 99.8% forecasting the highest accuracy. This research serves as a foundation for the decisions made by a variety of stakeholders, including analysts, policymakers, shareholders, and bank management, and it facilitates the comparison of the qualitative ratios of bankruptcy. The goal is to develop a prediction system that will allow the firms and businesses to be categorized according to the level of risk. 2025 Author(s). -
On the influencing facets of infant mortality in Karnataka: A study based on birth orders
The infant mortality rate (IMR) is used to assess the overall physical health of any community. Reducing this and spreading awareness among people can improve the well-being of society. In India, IMR is high due to the complex and challenging health policies and increased population, but various socio-economic and demographic factors play a significant role in determining the infant mortality rate. This study majorly focuses on identifying the factors influencing infant mortality, and a model has been proposed to estimate the likelihood of an infant's survival in Karnataka. For the empirical analysis, data has been taken from the National Family Health Survey-4 (2015-16), India. It is found that mothers' education and female literacy are the most significant factors affecting the IMR irrespective of the birth order. It is also found that the various socio-economic and demographic factors do not have a significant influence on the survival status of an infant as the birth order increases. Other factors like preceding birth interval, wealth index, caste, and religion also influence infant mortality. Hence, it is suggested that parents should have access to quality education and health facilities near their place of residence to reduce infant mortality at each order of birth. 2025 Author(s). -
Lightweight Sybil attack detection framework for wireless sensor network with cluster topology
The development of communication and networking technology has made it possible for wireless sensor networks to play a significant role in many fields. Wireless sensor networks are vulnerable to a variety of security threats because of their remote hostile features. The Sybil attack, which generates several identities to gain access to wireless sensor networks, is one such devastating but simple to spread exploit. This Paper proposes a novel identity and trust-based system to ensure protection against Sybil attacks. Analysis of the RSSI and location parameter increases the accuracy. It recognises the attackers and broadcasts information about them to all adjacent sensor nodes. Additionally, it offers other crucial security features. 2025 Author(s). -
Automated Door with Password-Based Lock
The application of this work is to lock the door and ensure the safety of our space. This was done with heavy locks earlier. Locks do not ensure safety completely and there is a lot of tension around them. The main problem with traditional locks is that they are heavy, and their system is completely mechanical. The three basic ideas of this project are safety, privacy, and automation. This device is a password-based door lock system in which the door is opened and closed without any physical work, i.e. automatically. The key here is the password that the user has to enter to open the door. When the correct password is entered into the keypad, the microcontroller gives a command to the servo motor to rotate at a specific angle. If the incorrect password is entered, the motor will not do any operation and the user will not be allowed to enter. 2025 Author(s).