Computer Engineering, Computer Science, Computer Networks and Communications, Computational Theory and Mathematics
32
Scopus Publications
Scopus Publications
Mitigating Cybersecurity Challenges in Smart City Development: Threats, Risks, and Strategic Solutions M. Vijayalakshmi, L. A. Anto Gracious, P. Girija AI Cybersecurity and Next Generation Mobility in Smart Cities, 2026 Smart cities are the result of significant innovations like IoT, AI, and 5G, and new forms of cyber threats have appeared. This chapter aims to identify and discuss the major cybersecurity threats, risks, and challenges in smart cities, such as IoT-based threats, privacy and data protection issues, and network-level threats. It also assesses the solutions that have been adopted at the strategic level with a view to managing these risks, including encryption, the use of artificial intelligence in threat detection, and blockchain in matters concerning data authenticity. The study focuses on the need to have policy structures in place, international cooperation, and constant development to meet emerging threats in the field of cybersecurity. As more and more smart cities are being developed, it will become paramount to ensure that there are adequate security measures put in place to safeguard the infrastructure that will be in place as well as the citizens.
IoT-integrated Water Quality Surveillance and Smart Feeding for High-Density Pisciculture Sustainability S. Akash, Sampurna Sahoo, M. Vijayalakshmi Computing Communication and Intelligence, 2026 High-density fish farming, particularly using the BioFloc system, demands continuous monitoring and automation to ensure optimal water quality and feeding efficiency. This paper introduces an IoT-based automated system that combines real-time water quality monitoring and intelligent feeding to promote sustainability in aquaculture. The system employs a NodeMCU ESP32 microcontroller to track essential water parameters such as pH (6.8–7.5), total dissolved solids (TDS) (300–600 ppm), turbidity (30–100 NTU), and temperature (25–28°C), sending data to a cloud dashboard (Blynk) for remote access. A self-operating feeding mechanism powered by a NodeMCU ESP8266 ensures precise and timely feed distribution, minimizing waste and optimizing fish growth. The system is fully solar-powered, supporting off-grid operation and improving energy efficiency. Experimental results validate the system’s ability to maintain stable water conditions while enhancing feeding accuracy and overall sustainability. This solution provides a cost-effective, scalable, and sustainable approach to modern aquaculture.
Navigating the Campus: A Technological Framework for Efficient Wayfinding Solutions Dharshini M, Divyadarshini K, Vijayalakshmi M 7th International Conference on Innovative Trends in Information Technology Icitiit 2026, 2026 Navigating large and complex academic campuses presents significant challenges for students, faculty, and visitors, particularly in institutions spanning over 100 acres with multilayered buildings and diverse facilities. Conventional navigation aids such as static maps and signboards provide limited support and fail to offer real-time, context-aware guidance. To address these limitations, this paper presents a GPS-integrated mobile campus navigation system designed to deliver accurate, real-time, and user-centric wayfinding support. The proposed system provides interactive maps, turn-by-turn navigation with improved positioning accuracy, facility search, and personalized routing for key campus locations such as classrooms, libraries, administrative offices, and recreational centers. Secure authentication using institutional credentials ensures controlled access, while inclusive design features such as multilingual support and accessibility-oriented navigation enhance usability for users with diverse needs. The system is implemented using a cross-platform mobile framework and evaluated through user-centric testing, demonstrating a significant reduction in orientation time and navigation-related confusion, particularly for first-year students and visitors. The results indicate that the proposed solution offers a scalable and effective approach to improving campus navigation and overall user experience in higher education environments.
Smart Intelli Secure-Based LPG and Fire Safety System with Automatic Door Control M. Vijayalakshmi, P.G. Sivaranjan, Krithick Balaji Ramesh Computing Communication and Intelligence, 2026 The increase in the use of LPG, both for domestic purposes and commercial gains, has led many persons to express fears over gas leakage and fire incidents. Though conventional means of safety help to a certain extent, such as installing gas detectors and smoke alarms, they don’t have a coordinated safety response system. This article suggests an innovation in smart safety systems with integrated LPG and fire hazard detection, which includes biometric-based access control, automatic door operation, and real-time monitoring. Advanced gas and flame sensors will be taken into account to continuously monitor the environment and sound alarms when any hazard exists. The door against which the operation is done will simply unlock so that it allows the evacuation or entrance of people such as the emergency responders. Biometric security, through fingerprints, only allows access to the property to authorized persons while the rest are kept at bay. An Arduino microcontroller is the power for the entire system; it is also IoT-enabled for sending out real-time alerts to users as well as emergency services, thus offering low response times and providing awareness of current situations. This may well be used in modern houses and business establishments.
Deep Learning for Clinical Decision Support Systems G. Premalatha, K. Dhivya, Morenikeji E. Coker, M. Vijayalakshmi Deep Learning Models Towards Health Informatics Management Foundations Challenges and Opportunities, 2026 At present, deep learning is a significant technology that is reshaping the healthcare sector, especially through the development of Clinical Decision Support Systems (CDSS). CDSS enhance the management of intricate medical data, supporting healthcare professionals in making prompt, precise, and evidence-driven decisions. Deep learning techniques distinguish themselves from traditional rule-based systems by their ability to learn and identify patterns from large and diverse datasets, such as clinical writing, medical images, and electronic health records, without the need for explicit programming. This chapter explores the application of deep learning techniques, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformer models, in diverse fields such as disease prediction, risk stratification, and diagnostic support. Although the potential is considerable, there are ongoing challenges related to model interpretability, data privacy, and the seamless integration of these tools into current clinical workflows. This chapter highlights the growing potential of deep learning to enhance decision-making, reduce errors, and improve patient outcomes in modern healthcare environments, based on a review of recent developments and current research.
EnsembleEdgeFusion: advancing semantic segmentation in microvascular decompression imaging with innovative ensemble techniques B. Dhiyanesh, M. Vijayalakshmi, P. Saranya, D. Viji Scientific Reports, 2025 Semantic segmentation involves an imminent part in the investigation of medical images, particularly in the domain of microvascular decompression, where publicly available datasets are scarce, and expert annotation is demanding. In response to this challenge, this study presents a meticulously curated dataset comprising 2003 RGB microvascular decompression images, each intricately paired with annotated masks. Extensive data preprocessing and augmentation strategies were employed to fortify the training dataset, enhancing the robustness of proposed deep learning model. Numerous up-to-date semantic segmentation approaches, including DeepLabv3+, U-Net, DilatedFastFCN with JPU, DANet, and a custom Vanilla architecture, were trained and evaluated using diverse performance metrics. Among these models, DeepLabv3 + emerged as a strong contender, notably excelling in F1 score. Innovatively, ensemble techniques, such as stacking and bagging, were introduced to further elevate segmentation performance. Bagging, notably with the Naïve Bayes approach, exhibited significant improvements, underscoring the potential of ensemble methods in medical image segmentation. The proposed EnsembleEdgeFusion technique exhibited superior loss reduction during training compared to DeepLabv3 + and achieved maximum Mean Intersection over Union (MIoU) scores of 77.73%, surpassing other models. Category-wise analysis affirmed its superiority in accurately delineating various categories within the test dataset.
WGAN-GP based Framework for High-Fidelity Synthetic Fetal MRI Image Generation Sam T James, M. Vijayalakshmi, Kristen Talukdar Proceedings of 5th International Conference on Soft Computing for Security Applications Icscsa 2025, 2025 A small number of well annotated, high resolution fetal MRI datasets represent a serious barrier to the evolution of AI models applicable to the prenatal diagnosis sphere. This paper suggests an approach that uses Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) to generate synthetic images that overcome this shortcoming. The gradient-penalty loss used in WGAN-GP makes the training more stable and allows improved image fidelity. Random noise vectors are transformed into 256 x 256 x 3 synthetic MRI slices of anatomically realistic fetal brains by our generator network. The training was conducted on a dataset of fetal MRI slices with spectral normalization, AdamW optimizer, and a 5:1 critic-generator in training ratio. The measure of the Structural Similarity Index (SSIM) and manual assessment by clinical experts indicate that the created images are of high diversity and anatomical accuracy, outperforming standard GANs and DCGAN baselines. The findings provided by this paper show that WGAN-GP may be utilized in augmenting datasets used to train diagnostic models. The paper also presents the limitations and future perspectives such as use of the attention-based and conditional GANs. The proposed work will create a privacy-saving pipeline to create synthetic fetal MRIs and solve major concerns associated with scarcity or diversity of images.
Loan approval system through customer segmentation using big data analytics and machine learning International Journal of Advanced Science and Technology, 2020
Emotion recognition and regulation using multi-model system International Journal of Advanced Science and Technology, 2020
Face recognition door unlock system A. Prasath, Aditya Kumar, Akanksha Yadav, Bhuvishri Acharya, Momin Tauseef International Journal of Innovative Technology and Exploring Engineering, 2019
