IAES International Journal of Artificial Intelligence (IJ-AI)
Not a member yet
1769 research outputs found
Sort by
Improved convolutional neural networks for aircraft type classification in remote sensing images
Get PDFWith the exponential growth of available data and computational power, deep convolutional neural networks (CNNs) have become as powerful tools for a wide range of applications, ranging from image classification to natural language processing. However, during last decade, remote sensing imagery has emerged as one of the most prominent areas in image processing. Variations in image resolution, size, aircraft types and complex backgrounds in remote sensing images challenge the aircraft classification task. This study proposes an effective aircraft classification model based on CNN architecture. The CNN network architecture is improved to achieve more accuracy rate and to avoid overfitting and underfitting problems. To validate the proposed model, a new public aircraft dataset called multi-type aircraft remote sensing images 2 (MTARSI2) has been used. Through an analysis of existing methodologies and experimental validation, the model shows the superior performance of the proposed CNN model in comparison to state-of-the-art deep learning approaches
Comparative analysis of convolutional neural network architectures for poultry meat classification
Get PDFThe increasing demand for standardized food quality assurance, particularly in regions like Morocco, emphasizes the need for accurate classification of poultry meat. This study evaluates and compares ten convolutional neural network (CNN) architectures—VGG19, VGG16, ResNet50, GoogleNet, MobileNetV1, MobileNetV2, DenseNet, NasNet, EfficientNet, and AlexNet—for classifying commonly consumed poultry meat types in Moroccan markets, including chicken, turkey, fayoumi, and farmer’s chicken. A labeled image dataset was used to train and test each model, with performance assessed using metrics such as accuracy, precision, recall, training time, and computational complexity. Additionally, the study investigates how dataset size influences model performance, addressing challenges like limited data availability and scalability. The results highlight DenseNet as the top-performing architecture, achieving 98% classification accuracy while also demonstrating superior computational efficiency. These findings are valuable for improving food quality control, offering data-driven support for stakeholders in poultry production, distribution, and regulatory bodies. By identifying optimal deep learning models for poultry meat classification, the study contributes to enhancing food authentication and safety in Morocco and similar regions. It also encourages the integration of AI-driven systems in food inspection processes, providing scalable, accurate, and efficient solutions for ensuring standardized quality in the poultry supply chain
Hybrid convolutional vision transformer for extrusion-based 3D food-printing defect classification
Get PDFDeep learning is generally used to perform remote monitoring of three-dimensional (3D) printing results, including extrusion-based 3D food printing. One of the widely used deep learning algorithms for defect detection in 3D printing is the convolutional neural network (CNN). However, the process requires high computational costs and a large dataset. This research proposes the Con4ViT model, a hybrid model that combines the strengths of vision transformer with the inherent feature extraction capabilities of CNN. The locally extracted features in the CNN were merged using the transformers’ global features with four transformer encoder blocks. The proposed model has a smaller number of parameters compared to other lightweight pre-trained deep learning models such as VGG16, VGG19, EfficientNetB2, InceptionV3, and ResNet50. Thus, the proposed model is simplified. Simulations were conducted to classify defect and non-defect images obtained from the printing results of a developed extrusion-based 3D food printing device. Simulation results showed that the model produced an accuracy of 95.43%, higher than the state-of-the-art techniques, i.e., VGG16, VGG19, MobileNetV2, EfficientNetB2, InceptionV3, and ResNet50, with accuracies of 77.88, 86.30, 82.95, 90.87, 84.62, and 93.83%, respectively. This research shows that the proposed Con4ViT model can be used for 3D food printing defect detection with high accuracy
Transforming campus mobility: the DigiSticker system in digital parking solutions
Get PDFUniversity digital parking systems have several benefits and solve many problems with traditional parking. Universities without a digital parking system face restricted parking, traffic congestion, inefficient space utilization, security issues, limited decision-making data, and diminished sustainability initiatives. This study paper discusses the benefits of digital parking systems and the drawbacks of traditional methods. By using technology to streamline university parking administration, the DigiSticker system offers an innovative solution. The DigiSticker system improves parking efficiency, convenience, and security for students, professors, and staff by delivering real-time parking information, assistance, and automated payments. This system has a user-friendly website and mobile app, fast registration, gate access management, security, user experience enhancement, and sustainability. Universities can improve student and staff parking experiences while improving parking management efficiency, security, and sustainability by meeting these requirements
Human sentiment analytics using multi model deep learning approach
Get PDFFor assessing human beings, the measurement of willpower and human emotions plays an important role because human beings are emotional creatures. Emotional analysis, also known as sentiment analysis, is the process of using natural language processing (NLP) and machine learning to determine the emotions expressed in text, speech, or other forms of communication. However, critical emotional analysis is limited to human interactions only. Human emotional artificial intelligence or Human sentimental analytics, a sub domain of NLP seeks to improve this understanding. The Present study develops a model using multi model deep learning approach which is capable of efficiently understanding human emotions and their intentions, closely mirroring human cognition. By extending emotional analysis beyond the traditional limits, this model will collect broad ranging data to uncover clear and hidden emotional details. The primary objective of this paper is to build highly effective model which provides in-depth insights into human emotions, leading to logical conclusions depending on all available factors and reasons. The necessary input data for the current study will be collected from audio-visual media covering a vast range of audio and visual samples
Melanoma classification using ensemble deep transfer learning
Get PDFMelanoma, a type of skin cancer, poses significant challenges in early detection and diagnosis. Several methods for early melanoma detection, including visual inspection and several machine learning models, face challenges with accuracy. To overcome these issues, deep learning has been widely adopted in various biomedical applications. In this work, we employ deep transfer learning methods to classify melanoma. Firstly, we collect publicly available datasets containing melanoma images, their corresponding ground truth for segmentation, and class labels. Subsequently, we perform data preprocessing, normalization, and label encoding to address issues of varied illumination, image noise, and data imbalance. Next, we conduct feature extraction utilizing the previously trained deep learning models, VGG, ResNet, InceptionResNet, and MobileNet. The characteristic vectors obtained from each model are fused to produce a comprehensive depiction among the provided pictures. In the classification stage, we employ ensemble learning using transfer learning models, including EfficientNet, Xception, and DenseNet. These models are trained on the final feature vector to classify melanoma images effectively. The effectiveness of the suggested method is verified using publicly available ISIC 2017–2020 datasets, these model reports average accuracy scores of 96.10%, 97.23%, 97.50%, 98.33%, and 98.60%, in that order
Classification of Tasikmalaya batik motifs using convolutional neural networks
Get PDFThis paper presents a study on the classification of traditional Tasikmalaya batik motifs using convolutional neural networks (CNN). The experiments revealed that the high complexity of batik motifs significantly impacted model performance, as the handling of each class influenced the overall results. Initial experiments with the original dataset demonstrated suboptimal performance, characterized by accuracy and validation curves indicating overfitting, with only 75% accuracy achieved at a learning rate of 0.001, a batch size of 32, and 50 epochs. To enhance performance, we implemented data segmentation, data augmentation, optimized the choice of the best optimizer, utilized an optimal architecture, and conducted hyperparameter tuning. The best-performing model was trained on data subjected to specific preprocessing for each class, using the Adam optimizer with hyperparameter tuning set to a learning rate of 0.001, a batch size of 32, and 50 epochs. In the hyperparameter tuning experiment with the visual geometry group network (VGGNet) architecture, it was shown that there is an improvement in the prediction of the kumeli class, achieving an accuracy of 100%
Impact of smoothing techniques for text classification: implementation in hidden Markov model
Get PDFA hidden Markov model (HMM) is widely used for sequence modeling in various text classification tasks. This study investigates the impact of different smoothing techniques, such as Laplace, absolute discounting, and Gibbs sampling on HMM performance across three distinct domains: e-commerce products, spam filtering, and occupational data mining. Through the comparative analysis, Laplace smoothing consistently outperforms other techniques in handling zero-probability issues, demonstrating superior performance in the e-commerce and SMS spam datasets. The HMM without any smoothing technique achieved the best results for job title classification. This divergence underscores the dataset-specific nature of smoothing requirements, where the simplicity of parameter estimation proves effective in contexts characterized by a limited and repetitive vocabulary. Hence, the findings suggest that tailored smoothing strategies are crucial for optimizing HMM performance in different textual analysis applications
Securing post-quantum cryptography: side-channel resilience in CRYSTALS-Kyber key encapsulation mechanism
Get PDFThis study evaluates side-channel vulnerabilities in hardware implementations of the cryptographic suite for Algebraic lattices (CRYSTALS)-Kyber key encapsulation mechanism (KEM) using correlation and differential power analysis (DPA) techniques. Unprotected field-programmable gate array (FPGA) implementations across all Kyber parameter sets were successfully compromised, revealing significant information leakage. Attack complexity scaled linearly with key size. Additive Boolean masking provided varying protection levels, with 4-bit masking offering a 100× security increase at notable performance cost. Performance characterization showed increased slice utilization and reduced maximum frequency for higher-order masking. A novel hybrid countermeasure combining higher-order masking with controlled time randomization enhanced protection against machine learning-based attacks. Comprehensive power trace analysis using 12-bit precision at 500 MS/s sampling rates was conducted. Statistical evaluation utilized Pearson's correlation and Welch's t-tests with a 0.8 threshold for key recovery. Real world validation in IoT, financial, and satellite scenarios highlighted practical post-quantum cryptography (PQC) deployment challenges. The study provides concrete design guidance for efficiently securing hardware Kyber implementations against side-channel attacks
Dynamic service-aware network selection framework for multi objective optimization in 5G-advanced heterogeneous wireless networks
Get PDFThe increasing complexity of heterogeneous wireless networks (HWNs) and the diverse requirements of mobility patterns and service classes necessitate advanced solutions for network selection and resource optimization. Existing models often fall short in addressing dynamic mobility scenarios and service differentiation, leading to inefficiencies in resource allocation, suboptimal throughput, and increased latency. To overcome these limitations, this study proposes a dynamic service-aware network selector (DSANS) framework for 5G-advanced environments. The framework integrates an adaptive deep decision network (ADDN) for multi-objective optimization, addressing critical quality of service (QoS) metrics such as throughput, delay, and energy efficiency while enhancing quality of experience (QoE) for applications like enhanced mobile broadband (eMBB), ultra-reliable low latency communication (URLLC), and internet of things (IoT). The DSANS framework dynamically adapts to mobility patterns and varying network conditions, ensuring efficient resource estimation and optimal network selection. Simulation results highlight its superiority, achieving up to 25% improvement in throughput and a 15% reduction in latency compared to state-of-the-art algorithms. These findings validate DSANS as a robust solution for mitigating the limitations of existing models, optimizing network performance, and meeting the stringent demands of next-generation HWNs