Authors: Viraj P. Tathavadekar

Abstract: This study investigates the use of generative AI and human-centered design for sustainable solutions to the ever-more pervasive problems within software development processes with a particular view to improving cross-functional collaboration. The modeling of modern Software Development Life Cycles (SDLCs) is further complicated by such things as vague requirements, continuous changes, and integration problems, all of which delay projects and increase their cost. The enormous integration gaps that the present study identifies are connecting AI-driven technologies with human-centered design practices, especially in creating collaboration among varied teams but ensuring technology sustainability. Research objectives consist of studying generative AI's role in enhancing requirements gathering, design processes, and further automated intelligent decision-making in testing, application deployment, and maintenance. It also aims at understanding the challenges key stakeholders confront across different SDLC phases. This is being done through a mixed-methods research approach combining quantitative data on AI tool effectiveness: reducing technical debt and increasing efficiency in teams, with correspondent qualitative insights from industry case studies. The major outcomes indicate that AI-driven tools do not just improve the efficiency of processes, but are also conducive to sustainable development practice as they reduce resource consumption, promote better collaboration. Implications are that generative AI and human-centered design can transform SDLC practices leading to higher-quality products and much lower maintenance costs as well as overall sustainability in software development projects.

Published by: vikaspatanker

Authors: Goswami Jaygiri

Abstract: With the recent explosion of technology, cybersecurity is now a necessity to protect sensitive data, critical systems, and individual privacy. This review paper examines the existing state of cybersecurity, detailing some of the principal threats and countermeasures, related problems, and future paths. We particularly concentrate on AI, zero-trust architecture, blockchain, and quantum-resilient cryptography from a security viewpoint. We also touch on human factors, governance, and other cyber risk avoidance measures that deserve more research. This article provides a comprehensive overview of present and future directions in cybersecurity while assisting scholars, decision makers, and practitioners of cybersecurity.

DOI: 10.61137/ijsret.vol.11.issue3.116

Published by: vikaspatanker

Authors: Dinnesh Gr, Manoj Jai Sudhan, Mrs. A. Jeyanthi, Mrs.G.Priyaa Sri

Abstract: Stroke is a major global health challenge, contributing significantly to mortality and disability, and placing a heavy burden on healthcare systems. Timely and accurate diagnosis is critical to mitigate long-term complications and improve patient outcomes. This study introduces a hybrid deep learning framework for automated stroke detection in brain CT images, integrating Vision Transformer (ViT), LASSO regression, and DenseNet121 to enhance diagnostic accuracy and efficiency. Utilizing a Kaggle dataset of 1900 CT images (950 stroke, 950 normal), the system employs preprocessing techniques, including resizing to 224×224 pixels, grayscale-to-RGB conversion, and data augmentation (flipping, rotation, blurring), to ensure model robustness and adaptability. The ViT model extracts high-level semantic features, capturing global dependencies through self-attention mechanisms, which are then refined using LASSO regression for feature selection to reduce dimensionality and prevent overfitting. The refined features are fed into DenseNet121, a convolutional neural network optimized for efficient parameter usage and gradient flow, for binary classification (stroke vs. normal). A Tkinter-based graphical user interface facilitates seamless interaction, allowing radiologists to upload images and receive real-time predictions, enhancing clinical workflows. The system is designed for scalability, local deployment, and integration with hospital systems like PACS, addressing challenges of diagnostic delays and inter-observer variability. Evaluation on the dataset demonstrates robust performance, with an accuracy of 92.69%, precision of 91.36%, recall of 94.03%, and F1-score of 92.68%. These metrics underscore the system’s reliability in minimizing false negatives, critical for clinical applications. This framework advances automated stroke diagnosis by combining transformer and convolutional architectures, offering a scalable, interpretable solution for emergency settings and laying the groundwork for future enhancements in multi-class stroke classification and real-time deployment.

 

 

Published by: vikaspatanker

Authors: P.K.C.Ashara Rathnasinghe

Abstract: This document derives an analysis of factors influencing on digital banking adaptation by senior citizen in public license commercial banks in Colombo district and to know the dimensions driven by solutions for digital banking adaptation and the associated value proposition for senior citizen customers. The convenient function of digital banking has replaced interactions with physical money and reduced transaction time, better meeting the convenience needs in modern society life styles with the technological development. As digital banking concept plays an important role in day-to-day functions, understanding the factors which attracting consumers of senior citizen category by their age to use digital banking method will bring more opportunities for development, and further significantly improve the output in convenient way. This study discusses how to further influence the factors of digital banking adaptation by senior citizens who use public license commercial banks in Colombo district. This is based on the main theoretical framework of the selected 5 factors from several factors. In this study, data analysis is implemented by for the purpose of verifying the research model and hypotheses. The research results show that factors such as awareness of the service, lack of knowledge / training, cost of service, online security, perceived ease of use have selected as independent variables influence on senior citizens (age above 60 years) to adapt to use digital banking concept for their financial transactions. Three hundred and eighty-five number of samples will plan to select to the study and sample was consisted of random sampling technique. Statistical analysis and Regression analysis going to be used to confirm the impact of these five factors on digital bank adaptation.

Published by: vikaspatanker

Authors: P.K.C.Ashara Rathnasinghe

Abstract: This document derives an analysis of factors influencing on digital banking adaptation by senior citizen in public license commercial banks in Colombo district and to know the dimensions driven by solutions for digital banking adaptation and the associated value proposition for senior citizen customers. The convenient function of digital banking has replaced interactions with physical money and reduced transaction time, better meeting the convenience needs in modern society life styles with the technological development. As digital banking concept plays an important role in day-to-day functions, understanding the factors which attracting consumers of senior citizen category by their age to use digital banking method will bring more opportunities for development, and further significantly improve the output in convenient way. This study discusses how to further influence the factors of digital banking adaptation by senior citizens who use public license commercial banks in Colombo district. This is based on the main theoretical framework of the selected 5 factors from several factors. In this study, data analysis is implemented by for the purpose of verifying the research model and hypotheses. The research results show that factors such as awareness of the service, lack of knowledge / training, cost of service, online security, perceived ease of use have selected as independent variables influence on senior citizens (age above 60 years) to adapt to use digital banking concept for their financial transactions. Three hundred and eighty-five number of samples will plan to select to the study and sample was consisted of random sampling technique. Statistical analysis and Regression analysis going to be used to confirm the impact of these five factors on digital bank adaptation.

Published by: vikaspatanker

Authors: Jamila J, M. Sathya, R. Priyadharshini, J. Jenshya, V. Vinothini

Abstract: Urban areas face growing challenges in managing parking efficiently due to increased vehicle density. This paper proposes a real-time parking occupancy detection system using OpenCV, a powerful open- source computer vision library. By analyzing live video feeds from strategically positioned cameras, the system detects, classifies, and tracks vehicles using advanced object detection techniques, such as YOLO and SSD. This enables continuous monitoring of parking spaces and accurate assessment of their occupancy status. The system operates in three core stages: vehicle detection, classification of parked or moving vehicles, and real-time tracking using algorithms like Kalman filters or optical flow. Occupancy data is dynamically updated and shared via user-friendly interfaces such as mobile apps or digital displays, helping drivers find available spots efficiently. The system is developed using Python and OpenCV to ensure flexibility and ease of deployment across different parking environments. Performance evaluation was carried out using real-world datasets under various lighting and environmental conditions, demonstrating high accuracy and responsiveness. The proposed solution is scalable, adaptable to various camera setups, and suitable for deployment in street parking, garages, and smart city infrastructure. By improving parking space utilization and reducing the time spent searching for parking, this system contributes to easing traffic congestion, reducing fuel consumption, and enhancing the urban driving experience. With potential features such as safety compliance monitoring and modular architecture, the proposed system represents a significant step toward intelligent and efficient parking management in modern cities.

 

 

Published by: vikaspatanker

Authors: Ajay Singh, Om Ahire, Aditya Marathe, Jay Modiya, Aniket Gaikwad, Utkarsh Musale, Prof. Rajkumar Patil, Prof. Jyoti Nandimath

Abstract: In the fields of optimization and natural language processing (NLP), recent advances have introduced transformative methodologies that address complex challenges involving constraints and knowledge integration. Optimization under constraints, a crucial area of study, has been significantly enhanced by the use of asymmetric entropy measures. These measures provide a structured framework for solving boundary-specific problems, particularly in computational mathematics, by focusing on the interplay between statistical emulation, classification, and optimization techniques. Such approaches are particularly effective when solutions are dependent on hidden or undefined conditions, showcasing their utility in practical domains like environmental modeling and decision systems. In parallel, NLP has witnessed a dramatic evolution, with the emergence of frameworks like Retrieval-Augmented Generation (RAG), which integrates retrieval systems with generative models. This hybrid approach addresses the limitations of standalone generative models by providing contextually accurate and relevant responses. RAG has proven especially valuable for knowledge-intensive tasks, such as real-time question answering and complex decision-making, where the combination of retrieved factual data and generative capabilities creates outputs that are both precise and comprehensive. The ability of RAG to leverage live data sources further ensures that its outputs remain up-to-date, addressing the persistent issue of knowledge drift in AI systems. Furthermore, transformer-based architectures, including BERT and GPT, have redefined the paradigms of language understanding and generation. BERT’s bidirectional pre-training allows for an in-depth contextual comprehension of text, enabling it to excel in tasks such as sentiment analysis, entity recognition, and text classification. Meanwhile, GPT’s autoregressive nature focuses on generating coherent and contextually relevant text, making it ideal for applications requiring fluent language generation, such as conversational AI and creative content development. Advanced fine-tuning techniques, such as those applied in models like RoBERTa, have further enhanced the capabilities of these transformers by optimizing training processes and adapting them to domain-specific challenges, such as healthcare and legal analysis. The convergence of these fields has profound implications for real-world applications. By combining the structured decision-making frameworks of constrained optimization with the adaptive and context-aware capabilities of NLP models, researchers can address challenges that demand both precision and flexibility. For instance, in healthcare, this integration can enable AI systems to deliver accurate diagnoses and tailored recommendations by retrieving relevant medical knowledge and synthesizing it into user-friendly explanations. Similarly, in environmental modeling, the application of optimization techniques alongside NLP-driven data interpretation can enhance predictive capabilities and decision support systems. As these methodologies continue to evolve, their synergy opens new avenues for innovation. The seamless integration of optimization techniques, such as entropy-based frameworks, with transformer-based architectures not only improves performance but also ensures scalability across diverse domains. Applications in education, personalized recommendation systems, and automated content generation further illustrate the transformative potential of these combined approaches. This paper explores these intersections, proposing novel frameworks that leverage the strengths of both constrained optimization and advanced NLP techniques to deliver scalable, efficient, and contextually rich solutions for complex, real-world challenges.

 

 

Published by: vikaspatanker

Authors: Ajay Singh, Aditya Marathe, Aniket Gaikwad, Om ahire, Jay modiya, Utkarsh musale

 

Abstract: Artificial Intelligence (AI) continues to play a transformative role in healthcare, particularly through advancements in large language models (LLMs) and computer vision (CV). These technologies are now being increasingly applied in predictive healthcare systems to improve diagnosis, reduce human error, and enhance patient engagement. However, general-purpose pre-trained models often underperform in specialized medical contexts where accuracy, domain-specific knowledge, and multimodal understanding are essential. This research proposes a hybrid AI framework that combines natural language processing (NLP) and computer vision to support predictive and interactive healthcare use cases. In the NLP component of the system, we perform a comparative evaluation of six leading open-source LLMs—Mistral, FLAN-T5, GPT-Neo, DialoGPT, LLaMA, and Ollama—analyzing their adaptability to domain-specific tasks such as symptom triage, patient education, and medical question answering. These models were fine-tuned using full parameter updates and reinforcement learning from human feedback (RLHF), which allowed the models to better align their outputs with the nuanced communication styles and ethical expectations in clinical settings. In parallel, the CV module addresses a critical real-world challenge: automated prescription handwriting recognition, which is essential for minimizing misinterpretation of medication names and dosages. To tackle the variability and complexity of handwritten medical prescriptions, we utilize convolutional neural networks—specifically VGG16 and EfficientNet—for image-based classification and text recognition. A custom dataset of handwritten prescription images was created and annotated using domain knowledge, and the models were trained to map image inputs to structured medicine names. Our experiments reveal that EfficientNet, with its compound scaling and optimized architecture, outperforms VGG16 in both accuracy and training efficiency, particularly under noisy or low-resolution input conditions. By integrating these two components, we build a multimodal chatbot capable of receiving an image of a handwritten prescription, recognizing the medication using a CNN model, and generating an informative or advisory response using an LLM fine-tuned for medical NLP. This enables seamless user interaction, allowing patients or practitioners to interact with the system using both text and image inputs. Such a system has practical applications in telemedicine, hospital kiosks, pharmacy automation, and rural health outreach, where both human expertise and infrastructure may be limited. Our results demonstrate the effectiveness of combining LLM fine-tuning and CNN-based vision models for predictive healthcare. While larger LLMs like LLaMA and FLAN-T5 achieve higher accuracy in clinical language tasks, lighter models like DialoGPT and Mistral offer faster, more cost-effective deployment options. On the CV side, EfficientNet offers superior generalization with fewer parameters compared to legacy architectures. This research provides a comprehensive performance analysis and design framework for AI systems in healthcare, offering actionable insights into how different model configurations, training strategies, and hardware choices affect outcome quality and deployment feasibility.

DOI: http://doi.org/10.61137/ijsret.vol.11.issue3.119

 

Published by: vikaspatanker

Authors: Assistant Professor R. Ayyappan, Bavadharani C, Mehandhiga M, Gowtham K, Chandru J

 

 

Abstract: The Smart Gaze Wheelchair is an innovative assistive mobility system designed to empower individuals with physical disabilities, particularly those affected by paralysis, by enabling hands-free wheelchair navigation using facial gestures. This technology combines computer vision, sensor integration, and IoT-based monitoring to provide a comprehensive solution for mobility, safety, and health tracking. At the core of the system lies the ESP32 microcontroller, which processes input from a camera and sensors in real time. Eye gestures—such as blinking a specified number of times or turning the head in a particular direction—serve as intuitive controls for wheelchair movement. For example, users can start motion by turn on the switch, move forward by blinking three times, move backward with five blinks, and stop with six. Directional control is managed by turning the head left or right and blinking once, enabling precise navigation without the use of hands or physical exertion. To ensure the user’s well-being, the wheelchair is equipped with a pulse sensor that monitors heart rate and a DHT11 sensor that tracks environmental conditions such as temperature and humidity. These health parameters are transmitted in real time to the ThingSpeak IoT platform, allowing caregivers or medical staff to remotely monitor the user’s condition. Additionally, the system features an ultrasonic sensor for obstacle detection, preventing collisions, and an emergency SOS button that triggers instant alerts during distress. OpenCV, combined with a Sliding Window Algorithm, is utilized for facial gesture recognition, offering consistent performance even under variable lighting conditions. The wheelchair’s mobility is driven by DC motors connected to an L298N motor driver, ensuring smooth and responsive movement. This system not only improves the quality of life for users but also provides peace of mind to their families and caregivers. The Smart Gaze Wheelchair represents a significant step forward in accessible technology by combining independence, health monitoring, and safety into one cohesive and intelligent solution.

 

 

 

Published by: vikaspatanker

Authors: Dr. Jaspreet Kour, Tribhuti Kumar Gaurav, Utkarsh Trivedi, Utkarsh Singh

Abstract: This work proposes an IoT-based pacifier that tracks real-time critical infant health parameters, including body position, temperature, and respiratory rhythms. The pacifier is equipped with an Arduino microcontroller, ESP32 wireless module, thermostat for temperature measurement, an MPU6050 sensor for position tracking, and a microphone condenser for tracking breathing rates. Real-time transmission enables early abnormalities and quick notifications to the caregivers. With continuous health monitoring and prevention of disease hazards of respiratory distress and SIDS, this new system enhances infant security. Infants health monitoring is a vital aspect of neonatal care, which must be continuously monitored to detect abnormalities in its initial phase. The data are saved on a cloud server and retrieved through a particular mobile app, allowing caregivers to track baby health remotely. The system provides real-time warnings for abnormal motion or temperature sensing, allowing caregivers to intervene timely. The device has been designed to be small, non-invasive, and power-effective, allowing for infant comfort without sacrificing system reliability.

 

 

Published by: vikaspatanker