Patients with diabetes are at risk of developing a retinal disorder called Proliferative Diabetic Retinopathy (PDR). One of the main characteristics of PDR is the development of neovascularization, a condition in which abnormal blood vessels are formed on the retina. This condition can cause blindness if it is not detected and treated early. Numerous studies have proposed different image processing techniques for detecting neovascularization in fundus images. However, because of its random growth pattern and small size, neovascularization remains challenging to detect. Hence, deep learning techniques are becoming more prevalent in neovascularization identification because of their ability to perform automatic feature extraction on objects with complex features. In this paper, a method of neovascularization detection based on transfer learning is proposed. The performance of the transfer learning method is investigated using four pre-trained Convolutional Neural Network (CNN) models, which include AlexNet, GoogLeNet, ResNet18, and ResNet50. In addition, an improved network based on the combination of ResNet18 and GoogLeNet is proposed. Evaluation on 1174 retinal image patches showed that the proposed network could achieve 91.57%, 85.69%, 97.44%, and 97.10% of accuracy, sensitivity, specificity, and precision, respectively. We demonstrated that the proposed method outperforms each individual CNN for neovascularization detection. It also shows better performance compared to another method that utilized deep learning models for feature extraction and Support Vector Machine (SVM) for classification.

Proliferative Diabetic Retinopathy (PDR) is a severe retinal disease that threatens diabetic patients. It is characterized by neovascularization in the retina and the optic disk. PDR clinical features contain highly intense retinal neovascularization and fibrous spreads, leading to visual distortion if not controlled. Different image processing techniques have been proposed to detect and diagnose neovascularization from fundus images. Recently, deep learning methods are getting popular in neovascularization detection due to artificial intelligence advancement in biomedical image processing. This paper presents a semantic segmentation convolutional neural network architecture for neovascularization detection. First, image pre-processing steps were applied to enhance the fundus images. Then, the images were divided into small patches, forming a training set, a validation set, and a testing set. A semantic segmentation convolutional neural network was designed and trained to detect the neovascularization regions on the images. Finally, the network was tested using the testing set for performance evaluation. The proposed model is entirely automated in detecting and localizing neovascularization lesions, which is not possible with previously published methods. Evaluation results showed that the model could achieve accuracy, sensitivity, specificity, precision, Jaccard similarity, and Dice similarity of 0.9948, 0.8772, 0.9976, 0.8696, 0.7643, and 0.8466, respectively. We demonstrated that this model could outperform other convolutional neural network models in neovascularization detection.

This paper explores the application of deep learning models in waste material classification, motivated by the urgent need for efficient waste management practices to address environmental sustainability concerns. Drawing parallels with the success of deep learning in healthcare domains, the study investigates the effectiveness of various deep learning architectures for waste material classification. The DenseNet201 model is proposed and compared with various deep learning models such as ResNet, MobileNetV2, AlexNet, and GoogleNet. Experimental results demonstrate that DenseNet201 achieves superior accuracy, average recall, and average precision, making it the most effective model for waste material classification. The dense connectivity and feature aggregation capabilities of DenseNet201 contribute to its outstanding performance, showcasing its potential for enhancing waste management processes.

Traffic congestion in semiurban areas presents unique challenges due to limited road capacity, growing vehicle volumes, school zone peak demand, and environmental disruptions such as flooding. Traditional fixed-time traffic control systems often fail to adapt to dynamic traffic conditions, leading to prolonged delays, safety risks, and reduced mobility efficiency. This article investigates the potential of artificial intelligence (AI) and Internet of Things (IoT) technologies to improve traffic management in semiurban environments, with a case study focusing on school zones and flood-prone road segments. The proposed framework integrates AI-adaptive traffic signal control, IoT-based flood detection and road condition monitoring, and AI-driven school zone traffic management to enable real-time sensing, predictive decision making, and dynamic traffic optimization. By leveraging sensor networks, machine learning models, and privacy-preserving traffic concepts, the system enhances congestion mitigation, pedestrian safety, and operational responsiveness. Field observations and system design analysis indicate that AI- and IoT-enabled interventions may significantly reduce vehicle delays, improve traffic flow efficiency, and strengthen resilience against weather-related disruptions. The study highlights both the opportunities and practical deployment challenges of intelligent transportation systems (ITS) in semiurban settings, offering a scalable and cost-effective pathway toward smarter, safer, and more sustainable mobility.

The early detection of diabetic retinopathy (DR) in fundus images is crucial for preventing vision loss in diabetic patients. Deep learning models have played a significant role in advancing DR detection. This paper explores the relatively unexplored DarkNet19 model’s performance in comparison to well-established models like ResNet18 and ResNet50 for DR detection. A balanced dataset of healthy and DR images was created through standardization and augmentation techniques. The models underwent binary classification training and testing, and their performance was evaluated using accuracy and precision metrics. DarkNet19 outperformed the other models, achieving higher accuracy (0.7347) and precision (0.9233), demonstrating its potential to enhance early DR diagnosis and reduce the risk of vision loss. This research contributes to the field of DR detection, highlighting the effectiveness of DarkNet19.