Browse Items (16481 total)
Sort by:
-
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). -
Deep Learning-based Cybersecurity Framework for IoT Environments
The article "Cyber Security In IOT using deep learning Approach"presumably states the pressing need for increased cybersecurity within the rapidly growing Internet of Things (IoT) paradigm. It stresses the distinct problems brought by IoT settings, such as distributed systems and heterogeneous devices, that make the old methods of security inoperable. The article we underscores the promise of deep learning and AI as novel solutions for identifying and foiling cyberattacks, but also notes the emergence of adversarial AI employed by cybercriminals. In addition, it urges proactive cybersecurity measures and ongoing surveillance to counter changing threats, especially with IoT networks becoming increasingly complex with applications in smart cities and other industries. The abstract can conclude by emphasizing the need for continued research into AI-based cybersecurity solutions to guarantee effective protection against emerging threats. 2025 IEEE. -
Deep Learning-based Cybersecurity Framework for IoT Environments
The article "Cyber Security In IOT using deep learning Approach"presumably states the pressing need for increased cybersecurity within the rapidly growing Internet of Things (IoT) paradigm. It stresses the distinct problems brought by IoT settings, such as distributed systems and heterogeneous devices, that make the old methods of security inoperable. The article we underscores the promise of deep learning and AI as novel solutions for identifying and foiling cyberattacks, but also notes the emergence of adversarial AI employed by cybercriminals. In addition, it urges proactive cybersecurity measures and ongoing surveillance to counter changing threats, especially with IoT networks becoming increasingly complex with applications in smart cities and other industries. The abstract can conclude by emphasizing the need for continued research into AI-based cybersecurity solutions to guarantee effective protection against emerging threats. 2025 IEEE. -
Deep Learning-Based Consolidated Disease Classification in Health Data Management
Healthcare data management is critical for ensuring comprehensive and high-quality medical treatment. Sensitive patient data management has a potential new option thanks to blockchain technology. However, existing blockchain-based healthcare data management systems face challenges in scalability, integration, and regulatory compliance. To address these issues, a novel blockchain-based healthcare data management system has been proposed to provide a secure, decentralized, and interoperable platform for managing sensitive patients medical information. Proposed approach involves collecting comprehensive health measurements from patients using wearable sensors and ensuring the security and integrity of patient data through robust user verification protocols. Artificial Neural Networks (ANNs) are employed to consolidate disease symptoms, enhancing the efficiency and accuracy of data analysis. The results and comparative analysis showcase the efficiency of the proposed method in terms of precision, accuracy, recall, search accuracy and F1-score. The accuracy of the proposed method is improved by 12.9%, 6.07%, and 14.28% when related to the existing ACTION-EHR, BSDMF, and BlockMedCare techniques respectively. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025. -
Deep Learning-Based Approach for Automated Cataract Detection
Advancements in deep learning approaches is of profound significance in the early detection of cataracts. Automated cataract detection using deep learning approaches is proposed in this chapter. Initially, two pretrained custom convolutional neural network (CNN) architectures, VGG-19 and MobileNetV2, were implemented to detect cataracts. ODIR-5K dataset is used for training, testing, and validating these models, and it has almost 6,400 fundus images. This preprocessed dataset provides the metadata of the available images and is labeled with diagnostic keywords. Since the dataset is highly imbalanced, class weighting techniques are utilized to avoid the impact of the imbalanced dataset. The performance of the models is evaluated, and results show that the ensemble approach outperforms other pretrained models, demonstrating the efficacy of hybrid CNN architecture in enhancing the accuracy of the diagnosis process. 2026 selection and editorial matter, T. Ananth Kumar, R. Rajmohan, M. Niranjanamurthy and G. Sambasivam. -
Deep Learning Model with Enhanced Segmentation and Combined Feature Activation for Mitosis Classification
Mitosis is a cell division mechanism vital for the growth of tissues and repair, Histopathological images are used by pathologists to diagnose cancer, but mitosis classification plays an important role in disease diagnosis. The mitotic counts are a proliferative indicator to find the aggressiveness of breast cancer. Detecting the mitotic tumor cells in tissue areas is a critical marker in cancer prognosis. Various researchers have focused on developing an automatic detection framework to identify mitotic figures, but detecting and classifying mitosis accurately remains a significant challenge in the medical field. Moreover, this research has designed a proposed Aggressive Tracing Seeking Optimization (ATSO) based Deep Convolutional Neural Network (Deep CNN) for the mitosis classification framework. The proposed framework uses less memory and increases the convergence rate; hence, it is globally efficient in achieving optimal solutions in the search space. The inspiration for considering the ATSO is its excellent behavior, as well as its scalable and adaptable mechanism, which allows optimization to move away from local optima. Moreover, it is computationally faster and exhibits higher global convergence capability in searching for the best solution. ATSO optimally trains a Deep CNN to generate higher classification accuracy by minimizing the false rate using the loss function. More explicitly, the proposed ATSO-Deep CNN model attained higher performance with an accuracy of 96.31%, an F1-score of 96.3%, precision of 96.84%, and recall of 95.78% with a 90% training percentage for the BreCaHAD dataset. 2025 Inventive Research Organization. -
Deep Learning Inspired Nonlinear Classification Methodology for Handwritten Digits Recognition Using DSR Encoder
The overlapped handwritten digit classification is a global challenge and a significant measure to assess the network recognition ability ratio. Most efficient models have been designed based on convolutional neural networks (CNN) for effective image classification and digit identification. Subsequently, multiple CNN models have inadequate accuracy because of high degree parameter dimensions that lead to abnormal digit detection error rates and computation complexity. We propose a Deep Digit Recognition Network (DDRNet) based on Deep ConvNets to minimize the number of parameters and features to keep the model light while maximizing the accuracy with an adaptive voting (AV) scheme for digit recognition. The individual digit is identified by CNN, and uncertain digits or strings are identified by Deep Convolutional Network (DCN) with AV scheme through Voting-Weight Conditional Random Field (VWCRF) strategy. These methods originated with the YOLO algorithm. The simulations show that our DDRNet approach achieves an accuracy of 99.4% without error fluctuations, in a stable state with less than 15 epochs contrast with state-of-art approaches. Additionally, specific convolution techniques (SqueezeNet, batch normalization) and image augmentation techniques (dropout, back-propagation, and an optimum learning rate) were examined to assess the system performance based on MNIST dataset (available at: http://yann.lecun.com/exdb/mnist/). 2022, King Fahd University of Petroleum & Minerals. -
Deep Learning in Waste Management and Recycling in Digital Smart City
For waste management and recycling in smart cities, the fast growth of urban populations and the subsequent rise in garbage creation have posed considerable issues. For cities to be sustainable and ecologically friendly, good waste management and the promotion of recycling practises are crucial. Deep learning techniques have become a potent tool for solving complicated issues and streamlining numerous procedures in a variety of fields in recent years. In the framework of smart cities, this chapter proposes improved Deep learning model with IOT Architecture for recycling and garbage management. 2025 Scrivener Publishing LLC. -
Deep Learning in Project Planning and Scheduling
Controlling construction projects requires careful planning, and the most popular modelling techniques are the discrete-event simulator (DES), linear schedule (LS), and the critical path method (CPM). DES techniques, however, may become laborious and struggle to appropriately represent decision possibilities as complexity and restrictions increase. Through the reinforcement learning methods, deep learning-based artificial intelligence (AI) may be a viable substitute, enabling a quicker evaluation and suggestion of planning solutions for intricate building projects. This study investigates if artificial intelligence (AI) can replace DES in Insight, an illustrated constraint-based procedure planning tool for production and building. In the study, the difficulties of integrating AI into planning for building are discussed, along with the process modifications required to support deep learning techniques. Enhanced schedule, expenses, and efficiency in operation result from early planning of projects, which also balances conflicting project requirements. The planning of modern building projects is suggested to use a new conceptual methodology. 2025 IEEE. -
Deep learning framework for stock price prediction using long short-term memory
Forecasting stock prices is always considered as complicated process due to the dynamic and noisy characteristics of stock data influenced by external factors. For predicting the stock market, several approaches have been put forward. Many academics have successfully forecasted stock prices using soft computing models. Recently, there has been growing interest in applying deep learning techniques in combination with technical indicators to forecast stock prices, attracting attention from both investors and researchers. This paper focuses on developing a reliable model for anticipating future stock prices in one day advance using Long Short-Term Memory (LSTM). Three steps make up the suggested model. The approach begins with ten technical indicators computed from previous data as feature vectors. The second phase involves data normalization to scale the feature vectors. Finally, in the third phase, the LSTM model analyzes the closing price for the next day using the normalized characteristics as input. Two stock markets, NASDAQ and NYSE are chosen to evaluate the efficacy of the developed model. To demonstrate how effective the new model is in making predictions, its performance is compared to earlier models. Comparing the suggested model to other models, the findings revealed that it had a high level of prediction accuracy. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024. -
Deep Learning for Uncovering of Fraud: A Design for Automated Financial Protection
Leveraging the unparalleled adaptability and hierarchical feature stratification capabilities of deep learning, this study constructs a sophisticated framework for fraud detection, seamlessly integrating convolution and recurrent neural architectures with advanced anomaly detection algorithms to decode complex, nonlinear transactional patterns within heterogeneous financial datasets, thereby enabling real-time fraud identification while addressing pivotal challenges of algorithmic interpretability, adversarial resilience, regulatory compliance, scalability, and data confidentiality, ultimately redefining the paradigm of automated financial security in an increasingly digitized global economy. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026. -
Deep Learning for Sustainable Agriculture
The evolution of deep learning models, combined with with advances in the Internet of Things and sensor technology, has gained more importance for weather forecasting, plant disease detection, underground water detection, soil quality, crop condition monitoring, and many other issues in the field of agriculture. agriculture. Deep Learning for Sustainable Agriculture discusses topics such as the impactful role of deep learning during the analysis of sustainable agriculture data and how deep learning can help farmers make better decisions. It also considers the latest deep learning techniques for effective agriculture data management, as well as the standards established by international organizations in related fields. The book provides advanced students and professionals in agricultural science and engineering, geography, and geospatial technology science with an in-depth explanation of the relationship between agricultural inference and the decision-support amenities offered by an advanced mathematical evolutionary algorithm. 2022 Elsevier Inc. All rights reserved. -
Deep Learning for Sustainable Agriculture
The evolution of deep learning models, combined with with advances in the Internet of Things and sensor technology, has gained more importance for weather forecasting, plant disease detection, underground water detection, soil quality, crop condition monitoring, and many other issues in the field of agriculture. agriculture. Deep Learning for Sustainable Agriculture discusses topics such as the impactful role of deep learning during the analysis of sustainable agriculture data and how deep learning can help farmers make better decisions. It also considers the latest deep learning techniques for effective agriculture data management, as well as the standards established by international organizations in related fields. The book provides advanced students and professionals in agricultural science and engineering, geography, and geospatial technology science with an in-depth explanation of the relationship between agricultural inference and the decision-support amenities offered by an advanced mathematical evolutionary algorithm. 2022 Elsevier Inc. All rights reserved. -
Deep Learning for Stock Market Index Price Movement Forecasting Using Improved Technical Analysis
Equity market forecasting is difficult due to the high explosive nature of stock data and its impact on investor's stock investment and finance. The stock market serves as an indicator for forecasting the growth of the economy. Because of the nonlinear nature, it becomes a difficult job to predict the equity market. But the use of different methods of deep learning has become a vital source of prediction. These approaches employ time-series stock data for deep learning algorithm training and help to predict their future behavior. In this research, deep learning methods are evaluated on the India NIFTY 50 index, a benchmark Indian equity market, by performing a technical data augmentation approach. This paper presents a Recurrent Neural Network (RNN), Long Short Term Memory (LSTM), and the three variants of Gated Recurrent Unit (GRU) to analyze the model results. The proposed three GRU variants technique is evaluated on two sets of technical indicator datasets of the NIFTY 50 index (namely TA1 and TA2) and compared to the RNN and LSTM models. The experimental outcomes show that the GRU variant1 (GRU1) with TA1 provided the lowest value of Mean Square Error (MSE=0.023) and Root Mean Square Error (RMSE= 0.152) compared with existing methods. In conclusion, the NIFTY 50 index experiments with technical indicator datasetTA1 were more efficient by GRU. Hence, TA1 can be used to construct a robust predictive model in forecasting the stock index movements. 2021. All Rights Reserved. -
Deep Learning for Mental Health: Attention-Driven Multilayer CNN for Audio Depression Detection
Depressive Disorder is a common mental health problem that affects millions of people around the world. This study proposes a Self-attention based Multi-layer Convolutional Neural Network (CNN) model to perform enhanced depression detection from audio modality. The model employs a diverse array of filters, kernel sizes, and pooling strategies across multiple CNN layers to capture local features, while the attention mechanism prioritizes emotionally salient parts of the speech signal, such as regions of low energy and lengthened pauses by assigning higher weights. Measured against the RAVDESS and TESS emotional speech datasets, the method attains an F1 score of 0.81, an accuracy of 83% and ROC-AUC of 0.96 when using attention, beating the baseline CNN model, F1 score of 0.77 and 83% accuracy without attention. The results demonstrate the effectiveness of attention-enhanced architectures in detecting depressive cues from speech and support the feasibility of developing real-world, speech-based mental health screening tools. 2025 IEEE. -
Deep learning for intelligent transportation: A method to detect traffic violation
Smart transportation is being envisaged as an important parameter in building smart cities. Although conceptualized to have major advantages, lack of intelligent systems makes more vulnerable for disasters. The number of fatality due to road accident has increased up to 12% in 2022 as that of previous year says the WHO report. There are large number of new vehicles plying on roads which makes space constraint for the commuters. This makes a large number of traffic violations happening in urban areas. The smart cities insist and tries to adopt AI based methods for identifying traffic violations. Computer Vision are predominant solution in detecting traffic violation. This paper proposes a Deep learning method using famous YOLOV technique for object detection for effectively determining the traffic violation. The violations such as signal cross are concentrated in this research. The experimental results prove that the proposed technique has 95.1% of classification accuracy in detecting signal crosses. 2023 Author(s). -
Deep Learning for Early Detection of Tomato Leaf Diseases: A ResNet-18 Approach for Sustainable Agriculture
The paper explores the application of Convolutional Neural Networks (CNNs), specifically ResNet-18, in revolutionizing the identification of diseases in tomato crops. Facing threats from pathogens like Phytophthora infestans, timely disease detection is crucial for mitigating economic losses and ensuring food security. Traditionally, manual inspection and labour-intensive tests posed limitations, prompting a shift to CNNs for more efficient solutions. The study uses a well-organized dataset, employing data preprocessing techniques and ResNet-18 architecture. The model achieves remarkable results, with a 91% F1 score, indicating its proficiency in distinguishing healthy and unhealthy tomato leaves. Metrics such as accuracy, sensitivity, specificity, and a high AUC score on the ROC curve underscore the model's exceptional performance. The significance of this work lies in its practical applications for early disease detection in agriculture. The ResNet-18 model, with its high precision and specificity, presents a powerful tool for crop management, contributing to sustainable agriculture and global food security. (2024), (Science and Information Organization). All Rights Reserved. -
Deep Learning for Arrhythmia Classification: A Comparative Study on Different Deep Learning Models
Arrhythmias, or irregular heart rhythms, are a major global health concern. Since arrhythmias can cause fatal conditions like cardiac failure and strokes, they must be rapidly identified and treated. Traditional arrhythmia diagnostic techniques include manual electrocardiogram (ECG) image interpretation, which is time consuming and frequently required for expertise. This research automates and improves the identification of heart problems, with a focus on arrhythmias, by utilizing the capabilities of deep learning, an advanced machine learning technique that performs well at recognizing patterns in data. Specifically, we implement and compare Custom CNN, VGG19, and Inception V3 deep learning models, which classify ECG images into six categories, including normal heart rhythms and various types of arrhythmias. The VGG19 model excelled, achieving a training accuracy of 95.7% and a testing accuracy of 93.8%, showing the effectiveness of deep learning in the comprehensive diagnosis of heart diseases. 2023 IEEE. -
Deep learning ensembles for lung cancer detection in thoracic CT scans leveraging generative adversarial network technology
Effective treatment of lung cancer depends on early and accurate detection, which continues to be a major cause of cancer-related fatalities globally. Conventional diagnostic techniques are useful, but their efficacy in handling large amounts of thoracic computed tomography (CT) scan data is limited by their time-consuming nature and susceptibility to human error. The research here suggests a new deep learning model that integrates generative adversarial networks (GANs) for data improvement with a sophisticated ensemble approach to classification. GANs are employed to generate realistic synthetic CT images, addressing the challenges of limited datasets. The backbone of the proposed approach is a consensus-guided adaptive blending (CGAB) ensemble model that learns to dynamically combine the predictions of three top-performing convolutional neural networks (CNNs): ResNet-152, DenseNet-169, and EfficientNet-B7. The CGAB model improves prediction accuracy through model contribution weighting based on confidence scores and inter-model consensus, while a conflict-resolving auxiliary decision model is used. The approach was tested using the lung image database consortium and the image database resource initiative (LIDC-IDRI) dataset with a detection rate of 97.35, surpassing single-model and traditional ensemble methods. The current work provides a solid and scalable approach to lung cancer detection with better generalization, increased diagnostic consistency, and applicability for clinical use. This is an open access article under the CC BY-SA license. -
Deep Learning Ensemble for Neuro-Oncological Diagnostics: Fusion of VGG-16 and Convolutional Neural Network Architectures
Brain tumors are potentially fatal, prompt and accurate diagnosis is essential to appropriate treatment and management. MRI is a key method for locating tumors in the brain. This study introduces a HYBRID deep learning for binary classification of brain tumors, combining a pre trained VGG16 model with tailored CNN and Neural Networks. The fusion of these models is done via feature concatenation followed by a common classifier. This fusion helps in capturing both high-level abstract and task-specific features critical for classification. To help minimize overfitting and improve generalization, the models are subjected to rigorous data augmentation including rotation, zooming, and horizontal flipping, normalization, and resizing of images to 150 150 pixels. All models are trained and validated using the same data splits. Performance is determined by accuracy, training and validation loss, confusion matrices, and visualization with Matplotlib plots and Plotly which provide a vivid insight into the models. Experiments are conducted to determine the different model performances and the hybrid model attained an accuracy of 98.14%, which was higher than the standalone VGG16 (93%), CNN (91%), and NN (88%) models. 2025 IEEE.