Volume 13 | Issue 5 |

 Abstract

The accelerated proliferation of the Internet of Things (IoT) across industrial domains has elevated the need for secure and intelligent energy management frameworks. This study proposes an IoT architecture enhanced with blockchain for dynamic industrial energy optimization, supported by deep learning methodologies. Environmental and operational parameters such as temperature, electricity usage, and humidity are collected using sensors. These values are validated using SHA-256 hashing and securely stored on a decentralized Ethereum blockchain. A Deep Neural Network (DNN) model analyzes the processed data to identify trends and predict energy consumption behavior. Results are showcased via an interactive dashboard developed using React, enabling operators to proactively manage energy usage and detect anomalies. The integration of blockchain with AI ensures a transparent, scalable, and tamper-resistant solution for industrial energy monitoring. Experimentation confirms enhanced security, real-time anomaly detection, and reduced inefficiencies, establishing this approach as a resilient model for smart industry applications. Keywords: Blockchain, IoT, Deep Neural Network, Industrial Energy Monitoring, Data Integrity, Real-Time Analytics, Energy Forecasting.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Blockchain, IoT, Deep Neural Network, Industrial Energy Monitoring, Data Integrity, Real-Time Analytics, Energy Forecasting.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Deepfake technology has rapidly advanced in recent years, creating highly realistic fake videos that are difficult to distinguish from real ones. The increasing use of social media platforms and online forums has further complicated the detection of misinformation and malicious content. This study proposes a deep learning (DL)-based method for detecting deepfakes. The system comprises three components: preprocessing, detection, and prediction. Preprocessing includes frame extraction, face detection, alignment, and feature cropping. Convolutional neural networks (CNNs) are employed in the eye and nose feature detection phase, while a CNN combined with a vision transformer (ViT) is used for face detection. The prediction phase applies a majority voting mechanism to consolidate results from three models, each analyzing a different facial feature. The model is trained using FaceForensics++ and DFDC datasets. Performance is evaluated using accuracy, precision, recall, and F1-score, ensuring robustness and reliability against emerging deepfake threats.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Deepfake Detection, CNN, Vision Transformer, AI, FaceForensics++, DFDC, Feature Extraction, Majority Voting

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

BACKGROUND Scoring System were developed for the quantification and standardization of the severity of acute illnesses and for the prediction of outcome and to compare the severity of disease among patients in Intensive Care Units. These are commonly utilized in the ICU and are determined either upon admission or during the patient's stay. A number of physiological markers that have been shown to independently predict hospital death are assessed as part of the scores. Typically, Scoring Systems incorporate patient data that can be measured with ease into a statistical algorithm to provide a single score. These ratings indicate the course of a patient's disease or the way in which they will react to a therapeutic intervention. Most in critical care are related to the severity of the sickness, and some are made to forecast the outcome, which is usually death. In order to evaluate and track the results of interventions and therapies for use in research, traditional physiological Scoring Systems were first created. OBJECTIVES OF THE STUDY o To assess the level of Knowledge on Scoring System used in Critical Care Unit among the Nursing students of selected Colleges of Kamrup, Assam. o To find out the association of Knowledge on Scoring System used in Critical Care Unit among Nursing students of selected Colleges of Kamrup, Assam with the selected demographic variable. METHODS AND MATERIAL A descriptive research design was used in the study to accomplish the objectives. By using convenience sampling technique, sample size was selected. The study was conducted among 220 Nursing students of selected Colleges, Kamrup, Assam. Respondents were selected on the basis of the inclusion criteria. Semi- structured Knowledge questionnaire was used to assess the Knowledge. RESULTS The findings of the study showed that out of 220 Nursing students, majority of the Nursing students i.e. 95.9% belongs to the age group 20-25 years, 4.1% belongs to the age group of 26-31 years. Majority of the Nursing students i.e. 95.5% are female, whereas 4.5% are male. Majority of the Nursing students i.e. 59.5% are from BSc. Nursing 4th year and 40.5% from Post Basic BSc. Nursing 2nd year. 100% of the Nursing students are clinically exposed in ICU. 59.5% of the Nursing students are clinically exposed in ICU for 1 year and 40.5% of the Nursing students are clinically exposed in ICU for 4 months. 100% of the Nursing students have previous Knowledge on Scoring System used in Critical Care Unit. Majority of the Nursing students i.e 59.1% of the students have heard from hospital staff, 36.4% of the students have heard from teachers and 4.5% of the students have heard from workshop regarding Scoring System used in Critical Care Unit. Majority of the students i.e. 93.6% had moderately adequate Knowledge, 4.1% of the students had inadequate Knowledge and 2.3% had adequate Knowledge regarding Scoring System used in Critical Care Unit. The association was statistically tested for Knowledge with Fisher Exact test and analysis depicted that the demographic variable did not show statistically significant association with level of knowledge regarding Scoring System used in Critical Care Unit among Nursing students at p<0.05 level. CONCLUSION After analyzing data collected, 93.6% had moderately adequate Knowledge, 4.1% of the students had inadequate Knowledge and 2.3% had adequate Knowledge regarding Scoring System used in Critical Care Unit. Therefore, it is recommended that there should be more training programmes and awareness about the use of Scoring System in Critical Care Unit and implemented in the practice.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

KEYWORDS: Knowledge, Scoring System, Critical Care Unit, Nursing students.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Non-invasive glucose monitoring is a transformative approach to diabetes management, offering a painless and convenient alternative to traditional invasive methods. The GlucoSync Project, a novel system for non-invasive glucose prediction using the MAX30102 sensor and the Beer-Lambert Law. The system captures photoplethysmogram (PPG) signals at red and infrared wavelengths to calculate light absorbance, which is then used as input to a machine learning model for glucose prediction. Additionally, the system provides real-time heart rate and SpO2 measurements. The GlucoSync application features a secure, user- friendly interface, including login/logout functionality and a dynamic dashboard displaying users' health data. Experimental results demonstrate the feasibility of the approach, with the machine learning model achieving high accuracy in glucose prediction. GlucoSync represents a significant step toward low-cost, portable, and non-invasive glucose monitoring solutions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Non-invasive glucose monitoring, MAX30102 sensor, Beer-Lambert Law, Machine learning, Photoplethysmogram (PPG), Absorbance Rate, SpO2, Heart Rate.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Edge computing is an emerging computing paradigm that brings computation and data storage closer to the data source. Unlike traditional cloud computing, which relies on centralized infrastructure, edge computing performs data processing at or near the data generation point. This reduces latency, minimizes bandwidth usage, and improves system reliability. It plays a vital role in enabling real-time analytics, especially in applications like IoT, autonomous systems, smart cities, and healthcare. This paper reviews the foundational concepts, recent advancements, methodologies, system architectures, and technical challenges of edge computing. The integration of AI, 5G, lightweight OSs, containerization, and real-time response models have enhanced its real world applicability. The study also highlights security, privacy, and orchestration frameworks necessary for scalable and sustainable edge deployments.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Edge Computing, IoT, Real-Time Processing, Cloud-Edge Integration, AI at Edge, 5G, Edge Orchestration

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Brain tumor diagnosis using magnetic resonance imaging (MRI) is a critical step in clinical decision-making. In recent years, deep learning, particularly convolutional neural networks (CNNs), has revolutionized the landscape of automated medical image analysis. This review presents an extensive synthesis of state-of-the-art deep learning techniques employed in brain tumor classification using MRI. The discussion encompasses foundational CNN architectures, advanced models incorporating attention mechanisms, transformer-based designs, and emerging paradigms such as Capsule Networks and hybrid CNN-transformer frameworks. Emphasis is placed on understanding how model complexity, interpretability, and modality integration affect diagnostic performance. The review also covers key MRI modalities including T1, T2, FLAIR, and DWI, and their significance in accurate tumor delineation. Furthermore, it highlights major datasets, performance benchmarks, evaluation metrics, and clinical validation strategies. Challenges such as data scarcity, generalization, interpretability, and deployment constraints are thoroughly analyzed. Lastly, the paper outlines future directions including multimodal learning, explainable AI, and federated learning to bridge the gap between research prototypes and real-world clinical adoption. This review aims to guide researchers, clinicians, and developers in understanding the current landscape, identifying challenges, and exploring innovations that can lead to more robust and interpretable diagnostic systems for brain tumor classification.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Brain Tumor, MRI, Deep Learning, CNN, Classification, Medical Imaging

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

A brain stroke could be a possibly deadly neurological sickness that should be analyzed rapidly and precisely to lower the chance of passing or irreversible brain harm. Counterfeit intelligence's later presentation into the therapeutic industry has opened the entryway for inventive ways to bolster early stroke discovery. In arrange to make an viable and mechanized diagnostic tool, this investigate presents a Brain Stroke Location Framework based on CT looks utilizing profound learning calculations. To distinguish between cases of ordinary, ischemic, and hemorrhagic stroke, the framework primarily employments a Convolutional Neural Organize (CNN) engineering that has been prepared on a assortment of brain CT check datasets. To progress input quality and boost the model's flexibility, CT checks are preprocessed utilizing strategies such picture scaling, normalization, clamor decrease, and differentiate improvement. The profound learning demonstrate has been prepared to recognize minor designs and characteristics within the brain that human watchers would not see right absent. The framework shows tried and true execution in stroke classification when surveyed concurring to precision, affectability, and specificity. The innovation can significantly help radiologists within the early conclusion of strokes by giving solid and fast forecasts, which would abbreviate symptomatic times and empower incite restorative intercession. Through made strides clinical help, this activity seeks to improve by and large quiet results by progressing computer-aided diagnostic strategies in neurology.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Brain Stroke Detection, Deep Learning, CT Image, Convolutional Neural Network, Medical Imagining, Automated Diagnosis

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract: Property rights are crucial aspect of women's empowerment, economic independence, and social justice. Despite constitutional guarantees and legal reforms, women in India continue to face structural and societal barriers to owning and inheriting property. This paper examines the intersection of women's property rights, gender equality, and human rights within the Indian socio-legal framework. The study analyses constitutional provisions, statutory laws, and international instruments such as the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW). In these contexts, patriarchal customs, traditional inheritance practices, and a lack of legal awareness hinder women's access to land and property. Although enactments like the Hindu Succession (Amendment) Act, 2005 have aimed to address gender disparities in inheritance, their impact remains limited due to poor implementation and deep-rooted societal resistance. The Constitution of India, with its commitment to social and economic justice, offers a strong framework, but practical enforcement is lacking. Strengthening women's property rights is essential not only for upholding human rights but also for advancing the Sustainable Development Goals, reducing poverty, and promoting inclusive growth.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Women, Property Rights, Equality, Social Justice, Empowerment.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Accurate and timely classification of brain tumors is critical for effective clinical decision-making and treatment planning. This study presents a custom convolutional neural network (CNN) model developed for the automated classification of brain tumors using T1-weighted contrast-enhanced magnetic resonance imaging (MRI). The model categorizes images into four classes: glioma, meningioma, pituitary tumor, and no tumor. The model was trained over 10 epochs using the Adam optimizer and sparse categorical cross-entropy loss, achieving an overall test accuracy of 87%. Class-wise evaluation showed that the model performed best in identifying pituitary tumors and non-tumorous cases, with F1-scores of 0.92 and 0.90, respectively. Glioma detection yielded a high precision of 0.933, though recall was lower at 0.78, indicating potential under-detection. Meningioma classification was the most challenging, with an F1-score of 0.71, primarily due to visual similarity with other tumor types. The macro-averaged F1-score stood at 0.845, while the weighted average was 0.87, suggesting overall consistent and balanced model performance. The results confirm that a lightweight CNN can deliver reliable diagnostic support in classifying brain tumors, particularly for pituitary and no tumor categories. Nonetheless, further improvements--such as incorporating multimodal imaging and advanced data augmentation--are necessary to enhance the model's sensitivity and robustness, especially in complex cases. This work provides a foundation for scalable, interpretable AI solutions in neuro-oncology imaging.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Brain Tumor Classification, Convolutional Neural Network, MRI, Deep Learning, Medical Imaging, Diagnostic Support

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study has been conducted to analyse the consumers opinion and satisfaction level towards Jute products with special reference to Erode District. Jute fibres found useful in sophisticated fields like decorative and furnishing materials such as lamp shades, wall covers, curtains, upholsteries, etc. Today jute is the least expensive fibre of mass consumption and it is 100% bio-degradable, recyclable and environmental friendly. Jute is a natural fibre with golden and silky shine and hence called The Golden Fibre. It is now the second most important fibre in the world, next to cotton. This study has been conducted by using purposive random sampling method with a sample size of 100 respondents.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Importance of study, Role of Jute Products, Major problems in purchase of jute products

  License

Creative Commons Attribution 4.0 and The Open Definition