A modified YOLO-based approach for classification and detection of crop-weed in sesame crops Wael Hadi, Sandip Sonawane, Tushar Jaware, Tarek Khalifa, Nawaf Ali, Faisal Aburub Peerj Computer Science, 2026 Weed plants pose a major threat in modern agriculture as they vie with primary crops for vital resources. They contribute to higher agricultural expenditure and diminished farm productivity, thereby influencing global agricultural economy. This manuscript proposes a system for classification and detection of crop and weed in sesame crops. In this system, different versions of Convolutional Neural Networks-based You Only Look Once (YOLO) object detection methods have been modified and the performance of YOLOv5, YOLOv6, and YOLOv7 compared. The proposed work utilized two datasets: a public weed dataset and a custom (own-created) dataset. The public weed dataset comprises 1,300 images, while the custom dataset includes 2,148 real-time images. Our investigation demonstrates that the YOLOv5 algorithm outperforms YOLOv6 and YOLOv7 algorithms in terms of evaluation measures like mean average precisions (mAPs), precision and recall. The YOLO models, particularly YOLOv5, demonstrated notable promise for the identification of weed in sesame crops. The efficacy of proposed approach is compared with that of existing approaches.
BER and power consumption minimization through optimization in wireless cellular network Gajanan Uttam Patil, Anilkumar Dulichand Vishwakarma, Priti Subramanium, Tushar Hrishikesh Jaware International Journal of Informatics and Communication Technology, 2025 Quality of service (QoS) of wireless cellular networks affect due to more power consumption, maximum bit error rate (BER), minimum throughput and improper resource allocation. Improvement in QoS can be done by reducing power consumption, BER and enhancing throughput. Hence there is a need to address the approaches for reduction in power consumption, BER, enhancement in throughput and proper resource allocation through different schemes. In this paper grey wolf optimization (GWO) technique is investigated with different database functions and Its outcome is contrasted with alternative methods like particle swarm optimization (PSO) and genetic algorithm (GA), It is evident that the GWO algorithm performs exceptionally well in terms of BER and power consumption minimization than the other techniques. Hence the QoS of the wireless cellular network will not affect due to minimization of the BER and power consumption through our proposed scheme.
Enhancing Cardiovascular Disease Detection Through Exploratory Predictive Modeling Using DenseNet-Based Deep Learning Wael Hadi, Tushar Jaware, Tarek Khalifa, Faisal Aburub, Nawaf Ali, Rashmi Saini Computers, 2025 Cardiovascular Disease (CVD) remains the number one cause of morbidity and mortality, accounting for 17.9 million deaths every year. Precise and early diagnosis is therefore critical to the betterment of the patient’s outcomes and the many burdens that weigh on the healthcare systems. This work presents for the first time an innovative approach using the DenseNet architecture that allows for the automatic recognition of CVD from clinical data. The data is preprocessed and augmented, with a heterogeneous dataset of cardiovascular-related images like angiograms, echocardiograms, and magnetic resonance images used. Optimizing the deep features for robust model performance is conducted through fine-tuning a custom DenseNet architecture along with rigorous hyper parameter tuning and sophisticated strategies to handle class imbalance. The DenseNet model, after training, shows high accuracy, sensitivity, and specificity in the identification of CVD compared to baseline approaches. Apart from the quantitative measures, detailed visualizations are conducted to show that the model is able to localize and classify pathological areas within an image. The accuracy of the model was found to be 0.92, precision 0.91, and recall 0.95 for class 1, and an overall weighted average F1-score of 0.93, which establishes the efficacy of the model. There is great clinical applicability in this research in terms of accurate detection of CVD to provide time-interventional personalized treatments. This DenseNet-based approach advances the improvement on the diagnosis of CVD through state-of-the-art technology to be used by radiologists and clinicians. Future work, therefore, would probably focus on improving the model’s interpretability towards a broader population of patients and its generalization towards it, revolutionizing the diagnosis and management of CVD.
Intelligent Spectrum Access Control in Cognitive Radio Networks: A Q-Learning and MDP Approach Intelligent CR Anilkumar Dulichand Vishwakarma, Gajanan Uttam Patil, Tushar Hrishikesh Jaware, Priti Subramanium International Research Journal of Multidisciplinary Scope, 2025 Cognitive radio (CR) technology improves frequency resource usage through unlicensed users' opportunistic use of unused spectrum bands without disrupting licensed ones. With growth in wireless communication needs, dynamic spectrum access (DSA) has emerged as a fundamental concept in enhancing spectral efficiency. New CR systems are projected to outgrow traditional artificial intelligence (AI) models, adopting reconfigurable network infrastructures with the ability to manage autonomously elements to provide uninterrupted service quality. To aid this development, a metacognitive level providing self-monitoring learning and adaptation is necessary to fine-tune AI-based decisionmaking in real time. A new threshold optimization approach for cognitive radio networks, highlighting detection based on the Maximum-Minimum Eigenvalue (MME) criterion, is the theme of this work. The method combines Markov Decision Processes (MDPs) and Q-Learning to support smart spectrum allocation and adaptive spectrum sensing. By adaptively varying parameters based on feedback from the environment, the method enhances decision-making in uncertain and varying network conditions. Simulation outputs show that the model provides enhanced spectrum efficiency, shorter convergence time, and less interference, while maintaining Quality of Service (QoS) for secondary users. This research advances CR systems by marrying signal detection precision with smart learning paradigms to create the potential for strong, autonomous communication networks that can adapt to dynamic spectral conditions.
Automated Detection and Classification in Ovarian Disease Imaging Using YOLOv8 Model Tushar Hrishikesh Jaware, Gajanan Uttam Patil, Mayur D Jakhete, andip Ravindra Sonawane, Rashmi Saini International Research Journal of Multidisciplinary Scope, 2025 With over 324,000 new cases and over 200,000 fatalities recorded each year, ovarian cancer (OC), the seventh most frequent disease in women and the most deadly gynecological malignancy, is a major global health concern. Due to this increasing worldwide burden Cancer prevention is one of the biggest public health issues of the twenty-first century. Ovarian disease diagnosis presents unique challenges that demand efficient and accurate detection technique. An automated system for ovarian cancer identification and categorization is suggested to assist physicians. This study explores application of YOLOv8 model for detecting and classifying objects in ovarian disease imaging datasets. Utilizing a curated dataset with 3,518 images divided across training, testing, and validation sets, model was trained over 50 epochs with advanced augmentation techniques including Blur, CLAHE, and Median Blur. The training process achieved significant detection performance, yielding good precision (mAP@50) and mAP@50-95. Comprehensive evaluation revealed class-specific challenges, including imbalances and variations in detection precision and recall rates. The integration of Tensor board visualizations further supports detailed performance analysis. The findings demonstrate YOLOv8's potential in advancing automated diagnostic tools for ovarian disease research and suggest that YOLOv8 can be used to predict ovarian cancer. Offering insights into future improvements in model optimization and dataset enhancement for clinical applications.
Effective CT Lung Image Denoising using Deep-Dense Inception Generative Adversarial Network Narendra Lalchand Lokhande, Tushar Hrishikesh Jaware International Research Journal of Multidisciplinary Scope, 2025 Computed tomography (CT) is used to visualize body structures and diagnose anomalies, making it an important tool in medical diagnosis and therapy planning. However, imaging techniques such as CT, MRI, ultrasound (US), and PET are frequently hampered by numerous types of noise, including Gaussian, speckle, Poisson variability, and salt-andpepper disturbances. These noises are created by technological interference, image processing flaws, and patient movement, which reduce image clarity and conceal key diagnostic details. The major difficulty in medical imaging is to remove noise while retaining important diagnostic information. Traditional denoising algorithms, such as Gaussian, median, and Wiener filters, frequently fail to adequately control complicated noise patterns or preserve small image details, limiting their utility in medical applications. This study presents an advanced unsupervised blind image denoising strategy that use an integrated model to treat numerous noise types without requiring paired noisy and clean images. The suggested method uses a deep and dense generative adversarial network (DD-GAN) with a new loss function to efficiently reduce noise and degradation at various intensity levels. This method advances CT image denoising by tackling issues such as intra-class variability, artefact importance, and training complexity, hence enhancing diagnostic reliability and accuracy
WGAN-LUNet for High-Accuracy Lung Nodule Segmentation Narendra Lalchand Lokhande, Tushar Hrishikesh Jaware International Research Journal of Multidisciplinary Scope, 2024 In the realm of computer-aided diagnosis systems designed for lung cancer, accurately segmenting nodules holds vital importance. This segmentation process has a vital role in examining the image attributes of lung nodules captured in computed tomography scans, ultimately aiding in separation of benign and cancerous nodules. Timely detection of these lesions stands as the most effective strategy in combating lung cancer, a disease notorious for its high malignancy rates across both genders. Despite numerous deep learning techniques proposed for nodule segmentation, it remains challenging due to factors such as nodule characteristics, location, false positives, and the necessity for precise boundary detection. The present paper presents an ultra-modern method for lung nodule segmentation in computer tomographic images, based on a Generative Adversarial Network. A discriminator and a generator make up the GAN model. Our generator, Residual Dilated Attention Gate UNet, serves as the segmentation module, while a discriminator is Convolutional Neural Network classifier. To enhance training stability, we utilize the Wasserstein GAN algorithm. We compare our hybrid deep learning model, called WGAN-LUNet, both quantitatively and qualitatively with other methods that are already in use. We evaluate the model using multiple quantitative criteria.
Enhanced Neonatal Brain Tissue Analysis via Minimum Spanning Tree Segmentation and the Brier Score Coupled Classifier Tushar Hrishikesh Jaware, Chittaranjan Nayak, Priyadarsan Parida, Nawaf Ali, Yogesh Sharma, Wael Hadi Computers, 2024 Automatic assessment of brain regions in an MR image has emerged as a pivotal tool in advancing diagnosis and continual monitoring of neurological disorders through different phases of life. Nevertheless, current solutions often exhibit specificity to particular age groups, thereby constraining their utility in observing brain development from infancy to late adulthood. In our research, we introduce a novel approach for segmenting and classifying neonatal brain images. Our methodology capitalizes on minimum spanning tree (MST) segmentation employing the Manhattan distance, complemented by a shrunken centroid classifier empowered by the Brier score. This fusion enhances the accuracy of tissue classification, effectively addressing the complexities inherent in age-specific segmentation. Moreover, we propose a novel threshold estimation method utilizing the Brier score, further refining the classification process. The proposed approach yields a competitive Dice similarity index of 0.88 and a Jaccard index of 0.95. This approach marks a significant step toward neonatal brain tissue segmentation, showcasing the efficacy of our proposed methodology in comparison to the latest cutting-edge methods.
Preface Machine Learning for Mobile Communications, 2024
Machine learning for mobile communications Sinh Cong Lam, Chiranji Lal Chowdhary, Tushar Hrishikesh Jaware, Subrata Chowdhury Machine Learning for Mobile Communications, 2024
