Category Archives: Uncategorized

Authors: Ashwini Shinde, Dr. Kiran Wakchaure

Abstract: India’s agriculture sector remains the backbone of rural livelihoods and national food security, contributing substantially to economic growth and employment. However, farmers continue to encounter a wide range of structural and socio-economic barriers, including small and fragmented landholdings, heavy reliance on monsoon rains, inadequate technological adoption, post-harvest inefficiencies, financial vulnerabilities, and unstable market prices. Additional constraints such as rising labour expenses, low levels of mechanization, limited irrigation coverage, and insufficient knowledge of sustainable practices further limit agricultural productivity. This review paper explores these complex challenges in detail while assessing the effectiveness of major government programmes designed to address them. Key schemes—such as the Pradhan Mantri Fasal Bima Yojana (PMFBY), PM-Kisan income support, Soil Health Card initiative, e-NAM digital marketplace, Pradhan Mantri Krishi Sinchai Yojana (PMKSY), Minimum Support Price (MSP) mechanisms, and emerging digital agriculture efforts—are evaluated for their role in improving productivity, farmer income, and risk management. The study identifies notable policy successes as well as areas requiring improvement, emphasizing the need for integrated, technology-oriented, and farmer-focused strategies.

DOI: https://doi.org/10.5281/zenodo.18595876

Published by: vikaspatanker

Authors: Aliza Sayyad, Dr. Pravin Khatkale

Abstract: The prevalence of mental health conditions in- cluding stress, anxiety, and depression is on the rise worldwide, but stigma, ignorance, and a lack of mental health experts con- tinue to hinder early detection and ongoing emotional support. The Smart Mental Health Assistant, an AI-powered support sys- tem intended to assess user symptoms, forecast potential mental health issues, and offer tailored self-care advice, is the idea behind this project. The system incorporates a chatbot interface for user interaction and advice, Random Forest Classifier for mental health prediction, and Natural Language Processing (NLP) for symptom extraction. In order to effectively diagnose mental health disorders, the system transforms retrieved symptoms into binary vectors using datasets that include symptoms, severity lev- els, and preventative measures. This paper examines the body of research on AI in mental health, pinpoints important variables affecting technology uptake, and emphasizes the significance of scalable and easily accessible mental health resources. The re- sults show that early diagnosis, emotional monitoring, and pre- ventive treatments could all be enhanced by AI-based screening systems. By facilitating ongoing assistance, lowering stigma, and enhancing psychological well-being, the suggested assistant bene- fits the mental health ecosystem.

DOI: https://doi.org/10.5281/zenodo.18595836

Published by: vikaspatanker

Authors: Chinmay Chouhan, Assistant Professor Srashti Thakur

Abstract: Brain tumor segmentation is an important task in medical image processing. Early diagnosis of brain tumors plays an important role in improving treatment possibilities and increases the survival rate of the patients. Manual segmentation of the brain tumors for cancer diagnosis, from large amount of MRI images generated in clinical routine, is a difficult and time consuming task. There is a need for automatic brain tumor image segmentation. The purpose of this paper is to provide a review of MRI-based brain tumor segmentation methods. Recently, automatic segmentation using deep learning methods proved popular since these methods achieve the state-of-the-art results and can address this problem better than other methods. Deep learning methods can also enable efficient processing and objective evaluation of the large amounts of MRI-based image data. There are number of existing review papers, focusing on traditional methods for MRI-based brain tumor image segmentation. Different than others, in this paper, we focus on the recent trend of deep learning methods in this field. First, an introduction to brain tumors and methods for brain tumor segmentation is given. Then, the state-of-the-art algorithms with a focus on recent trend of deep learning methods are discussed. Finally, an assessment of the current state is presented and future developments to standardize MRI-based brain tumor segmentation methods into daily clinical routine are addressed.

Published by: vikaspatanker

Authors: Neha Sahu, Prof. Ruchika Saini

Abstract: This study focuses on the design and structural analysis of helical gears with varying helix angles to investigate their influence on mechanical performance. By designing helical gears with different helix angles and analyzing them under identical loading and boundary conditions, the study aims to evaluate changes in bending stress, contact stress, deformation, and axial force. The results of this investigation will help identify optimal helix angle ranges that enhance gear strength and longevity while minimizing undesirable effects such as excessive axial loads and material failure. The findings of this study are expected to contribute to improved gear design practices by providing insights into the relationship between helix angle variation and structural performance. Such insights are valuable for engineers and designers seeking to develop efficient, durable, and high-performance gear systems for modern mechanical applications.

Published by: vikaspatanker

Authors: Shuaib Ahmed

Abstract: The quick pace of AI coding assistant adoption in cloud engineering has greatly led to the creation of Infrastructure-as-Code (IaC) and CI/CD pipelines. Nevertheless, AI-generated setting may readily imply security misconfigurations, insecure defaults and violations of the policy that can be transmitted straight into production cloud environments. Such risks are especially acute in those organizations that deal with regulated and high-assurance industries, whose misconfigured resources can cause data breaches, privilege increases, and violation of the rules. Conventional security review procedures are too sluggish and manual to follow through with the AI-assisted development processes, which resulted in a pressing need of automated preventive security mechanisms. The paper presents a recommendation in the form of the AI Guardrailed Cloud Engineering Framework (AGCEF) that is a proactive security model that involves the imposition of guardrails on AI-generated IaC and CI/CD artifacts prior to the deployment. AGCEF combines policy-as-code checking, matching of vulnerability signatures, semantic intent checking with LLM and a quantitative risk scoring system, which identifies and thwart insecure configurations at design time. Through experimental analysis, it is shown that AGCEF is significantly better in comparison to current AI-based methods of vulnerability detection because it offers higher vulnerability prevention, lowers false negatives, less manual review, and enhances the safety of deployment. The framework allows organizations to use AI copilots to enhance productivity and maintain high levels of cloud security and compliance, hence restoring the balance between the speed of AI-assisted development and AI-assisted operations in the cloud.

DOI: https://doi.org/10.5281/zenodo.18594687

Published by: vikaspatanker

Authors: Mrs. M. Lavanya, Ms. R. Kavinila, Ms. M. Harini, Ms. K. Keerthana

Abstract: Education for students with disabilities continues to face challenges due to inadequate accessibility tools, lack of adaptive content, and poor connectivity in rural areas. Existing technologies such as screen readers, speech-to-text converters, and sign language translators function independently, resulting in fragmented learning experiences. EduAble, is a multi modal AI-powered learning platform designed to support students with visual, hearing, mobility, and neurodiverse challenges. It integrates Text-to-Speech (TTS), Speech-to-Text (STT), sign language and gesture recognition, and adaptive content simplification to create a unified, inclusive learning environment. EduAble is developed using Django with Django REST Framework for backend processing and React Native for cross-platform mobile accessibility, supported by PostgreSQL for data storage.The platform employs advanced AI models such as gTTS for speech synthesis, CNN with MediaPipe and OpenCV for gesture and sign language detection, and BERT for text simplification using TensorFlow which collectively enhance learning accessibility and provide a more effective and integrated assistive education system.

DOI: https://doi.org/10.5281/zenodo.18585544

Published by: vikaspatanker

Authors: Dr. Suman Dhawan

Abstract: Faculty Development Programs (FDPs) are very important to the careers of teachers in higher education institutions. They become better at what they do, and the entire institution is improved. But the fact is that faculty members are not all alike, and age is actually a factor in what they want from an FDP. This research explores how the interests of faculty change with age. On the basis of a structured survey of 302 faculty members from various universities and colleges, a One-Way ANOVA test was conducted to determine how needs differ in three age groups: 25-34, 35-44, and 45 and above. The findings are quite striking. Age does make a difference. The younger generation is more concerned with handling classes and establishing a sound foundation in subject matter. The middle-aged faculty begin to tilt towards competency development and professional growth. The 45+ age group is more concerned with developing their personalities and management acumen. To synthesise all this, this study proposes an Age-Life-Cycle Model of Faculty Development Priorities. The study concludes that "one-size-fits-all" solutions do not work. If universities are serious about faculty development, they need to listen to where people are in their life cycle and provide development that fits.

DOI: http://doi.org/10.5281/zenodo.18584290

Published by: vikaspatanker

Authors: Delos Santos, Greg, Galgo, Lady Nathalie A, Gubat, Karyll D., Zamora, Daisy Anne

Abstract: The purpose of this study is to design and develop a soil testing device known as TerraGrow using IoT technology that could help farmers test soil properties and recommend the appropriate crops for growing. The soil testing device could measure EC values, soil moisture levels, and soil temperatures to acquire valuable soil information, which could then be interpreted using a mobile or web application. The results obtained were analyzed using mean and percentage to test the accuracy of the soil testing device. The results revealed that the soil testing device TerraGrow could measure and interpret soil properties with greater accuracy and efficiency. The application of IoT technology made it easy for the soil testing device to store data and provide recommendations for growing appropriate crops based on soil quality. The study results showed that the soil testing device TerraGrow could work with greater efficiency and ease compared to traditional methods. The study concluded that the application of TerraGrow has made a significant contribution to modern agricultural practices. The soil testing device could be made even better with suggestions like automatic calibration and solar power.

DOI: https://doi.org/10.5281/zenodo.18539454

Published by: vikaspatanker

Authors: Sai Sukesh Reddy Tummuri

Abstract: Large language models powered by transformers have grown quickly, resulting in previously unheard-of performance improvements, but at the expense of high computational complexity, memory usage, and energy consumption. Their deployment in real-time and resource-constrained environments is hampered by these limitations. In order to improve inference efficiency while maintaining predictive accuracy, this paper proposed a novel Dynamic Sensitivity-Aware Quantization-Aware Training (DSA-QAT) framework. The suggested method adaptively adjusted quantization precision based on layer-wise sensitivity and training dynamics, in contrast to traditional quantization approaches that apply uniform precision reduction. This allowed for more informed precision allocation across transformer components. Using representative performance and efficiency metrics, controlled simulation experiments were used to assess the suggested framework. According to experimental results, the quantized model maintained balanced precision, recall, and F1-score values while achieving prediction accuracy above 97%. The model also demonstrated strong robustness against quantization noise, decreased inference latency, a smaller memory footprint, improved energy efficiency, and stable training loss convergence. Additionally, a notable decrease in model size was noted, allowing for effective deployment without sacrificing performance. Overall, the findings demonstrated that the suggested DSA-QAT framework successfully reduced the trade-off between accuracy and model efficiency. The study demonstrated the potential of adaptive quantization-aware strategies for the high-performance, scalable, and sustainable deployment of large language models in practical applications.

Published by: vikaspatanker

Authors: Riya Jaiswal, Simran Sahu, Prince Pandey, Vandana Thripathi

Abstract: Heart disease continues to be a major global health concern, accounting for a significant number of premature deaths each year. Early detection can improve survival rates, yet traditional diagnostic methods are time-consuming and often dependent on expert interpretation. This study applies machine learning techniques to clinical data to develop a predictive model capable of estimating heart disease risk. Various algorithms—including Logistic Regression, Random Forest, Support Vector Machine (SVM), and XGBoost—were evaluated. The results show that ensemble models deliver the highest accuracy, demonstrating strong potential for supporting clinical decision-making.

Published by: vikaspatanker