The agricultural sector plays a key role in supplying quality food and makes the greatest contribution to growing economies and populations. Plant disease may cause significant losses in food production and eradicate diversity in species. Early diagnosis of plant diseases using accurate or automatic detection techniques can enhance the quality of food production and minimize economic losses. In recent years, deep learning has brought tremendous improvements in the recognition accuracy of image classification and object detection systems. Hence, in this paper, we utilized convolutional neural network (CNN)-based pre-trained models for efficient plant disease identification. We focused on fine tuning the hyperparameters of popular pre-trained models, such as DenseNet-121, ResNet-50, VGG-16, and Inception V4. The experiments were carried out using the popular PlantVillage dataset, which has 54,305 image samples of different plant disease species in 38 classes. The performance of the model was evaluated through classification accuracy, sensitivity, specificity, and F1 score. A comparative analysis was also performed with similar state-of-the-art studies. The experiments proved that DenseNet-121 achieved 99.81% higher classification accuracy, which was superior to state-of-the-art models.

Diabetes mellitus is considered one of the main causes of death worldwide. If diabetes fails to be treated and diagnosed earlier, it can cause several other health problems, such as kidney disease, nerve disease, vision problems, and brain issues. Early detection of diabetes reduces healthcare costs and minimizes the chance of serious complications. In this work, we propose an e-diagnostic model for diabetes classification via a machine learning algorithm that can be executed on the Internet of Medical Things (IoMT). The study uses and analyses two benchmarking datasets, the PIMA Indian Diabetes Dataset (PIDD) and the Behavioral Risk Factor Surveillance System (BRFSS) diabetes dataset, to classify diabetes. The proposed model consists of the random oversampling method to balance the range of classes, the interquartile range technique-based outlier detection to eliminate outlier data, and the Boruta algorithm for selecting the optimal features from the datasets. The proposed approach considers ML algorithms such as random forest, gradient boosting models, light gradient boosting classifiers, and decision trees, as they are widely used classification algorithms for diabetes prediction. We evaluated all four ML algorithms via performance indicators such as accuracy, F1 score, recall, precision, and AUC-ROC. Comparative analysis of this model suggests that the random forest algorithm outperforms all the remaining classifiers, with the greatest accuracy of 92% on the BRFSS diabetes dataset and 94% accuracy on the PIDD dataset, which is greater than the 3% accuracy reported in existing research. This research is helpful for assisting diabetologists in developing accurate treatment regimens for patients who are diabetic.

Dysarthria, a motor speech disorder that impacts articulation and speech clarity, presents significant challenges for Automatic Speech Recognition (ASR) systems. This study proposes a groundbreaking approach to enhance the accuracy of Dysarthric Speech Recognition (DSR). A primary innovation lies in the integration of the SepFormer-Speech Enhancement Generative Adversarial Network (S-SEGAN), an advanced generative adversarial network tailored for Dysarthric Speech Enhancement (DSE), as a front-end processing stage for DSR systems. The S-SEGAN integrates SEGAN’s adversarial learning with SepFormer speech separation capabilities, demonstrating significant improvements in performance. Furthermore, a multistage transfer learning approach is employed to assess the DSR models for both word-level and sentence-level DSR. These DSR models are first trained on a large speech dataset (LibriSpeech) and then fine-tuned on dysarthric speech data (both isolated and augmented). Evaluations demonstrate significant DSR accuracy improvements in DSE integration. The Dysarthric Speech (DS)-baseline models (without DSE), Transformer and Conformer achieved Word Recognition Accuracy (WRA) percentages of 68.60% and 69.87%, respectively. The introduction of Hierarchical Attention Network (HAN) with the Transformer and Conformer architectures resulted in improved performance, with T-HAN achieving a WRA of 71.07% and C-HAN reaching 73%. The Transformer model with DSE + DSR for isolated words achieves a WRA of 73.40%, while that of the Conformer model reaches 74.33%. Notably, the T-HAN and C-HAN models with DSE + DSR demonstrate even more substantial enhancements, with WRAs of 75.73% and 76.87%, respectively. Augmenting words further boosts model performance, with the Transformer and Conformer models achieving WRAs of 76.47% and 79.20%, respectively. Remarkably, the T-HAN and C-HAN models with DSE + DSR and augmented words exhibit WRAs of 82.13% and 84.07%, respectively, with C-HAN displaying the highest performance among all proposed models.

Group management is practiced to deploy access control and to ease multicast and broadcast communication. However, the devices that constitute the Internet of Things (IoT) are resource-constrained, and the network of IoT is heterogeneous with variable topologies interconnected. Hence, to tackle heterogeneity, SDN-aided centralized group management as a service framework is proposed to provide a global network perspective and administration. Group management as a service includes a group key management function, which can be either centralized or decentralized. Decentralized approaches use complex cryptographic primitives, making centralized techniques the optimal option for the IoT ecosystem. It is also necessary to use a safe, scalable approach that addresses dynamic membership changes with minimal overhead to provide a centralized group key management service. A group key management strategy called a one-way Function Tree (OFT) was put forth to lower communication costs in sizable dynamic groups. The technique, however, is vulnerable to collusion attacks in which an appending and withdrawing device colludes and conspires to obtain unauthorized keys for an unauthorized timeline. Several collusion-deprived improvements to the OFT method are suggested; however, they come at an increased cost for both communication and computation. The Modified One-Way Function Tree (MOFT), a novel technique, is suggested in this proposed work. The collusion resistance of the proposed MOFT system was demonstrated via security analysis. According to performance studies, MOFT lowers communication costs when compared to the original OFT scheme. In comparison to the OFT’s collusion-deprived upgrades, the computation cost is smaller.

Skin Cancer classification has been one of the most challenging problems for dermatologists; it is a tremendously tedious process to detect the kind of lesion/cancer form it is for just the human eye. Deep learning has become popular due to its potential to learn complex traits from the huge dataset. A prominent deep learning model for image categorization is the convolutional neural network (CNN). Many researchers have been conducted on the efficiency of CNN

The academic and research communities are showing significant interest in the modern and highly promising technology of wireless mesh networks (WMNs) due to their low-cost deployment, self-configuration, self-organization, robustness, scalability, and reliable service coverage. Multicasting is a broadcast technique in which the communication is started by an individual user and is shared by one or multiple groups of destinations concurrently as one-to-many allotments. The multicasting protocols are focused on building accurate paths with proper channel optimization techniques. The forwarder nodes of the multicast protocol may behave with certain malicious characteristics, such as dropping packets, and delayed transmissions that cause heavy packet loss in the network. This leads to a reduced packet delivery ratio and throughput of the network. Hence, the forwarder node validation is critical for building a secure network. This research paper presents a secure forwarder selection between a sender and the batch of receivers by utilizing the node’s communication behavior. The parameters of the malicious nodes are analyzed using orthogonal projection and statistical methods to distinguish malicious node behaviors from normal node behaviors based on node actions. The protocol then validates the malicious behaviors and subsequently eliminates them from the forwarder selection process using secure path finding strategies, which lead to dynamic and scalable multicast mesh networks for communication.