SANKARA NARAYANAN S

@veltech.edu.in

Associate Professor/CSE
Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology

SANKARA NARAYANAN S


         Download Compact PDF  
https://researchid.co/sankar_bnec
21

Scopus Publications

Scopus Publications

  • Efficient IoT-Driven Smart Farming Framework Leveraging Multi-Scale Random Graph Diffusion Parallel Hybrid Networks for Accurate Crop Yield Prediction
    Kunal Devidas Gaikwad, S. Sankara Narayanan
    SN Computer Science, 2026
  • Automated Depression Diagnosis using EEG Signal Analysis
    Meena L C, Chidambarathanu K, Sankara Narayanan S, Karthikram Anbalagan
    Proceeding of International Conference on Computing Communication Control and Cyber Physical Systems I5cps 2026, 2026
    Depression, which is an endemic mental disorder, remains an enormous burden to the people and the society in general. Addressing this urgent issue, the given research attempts to offer the progressive solution to the problem of detecting depressive symptoms by examining the electroencephalogram (EEG) biomarkers. Using the potentials of the MNE library in Python, this paper will investigate a new approach of detecting signs of depression using EEG recordings. The data set used is EEG records of 181 patients in a hospital. Bandpass filter is employed to improve signal quality and minimize noise. The key characteristics such as Alpha, Beta and Theta band powers and such indices as Alpha Power Variation, Relative Gamma Wave and Spectral Asymmetry Index are then derived out of the preprocessed EEG data. These characteristics act as input variables to the training of four different models, namely, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree and Logistic Regression. These models are thoroughly tested using a supervised learning paradigm to determine the accuracy of the model in detecting depression when offered a similar dataset of EEG signals. The broader scope of this method would be to not only find EEG biomarkers in depression but also develop predictive models that can be reliable enough to differentiate individuals with depressive symptoms at an early stage. By combining data analysis of EEGs and supervised learning techniques, the study can contribute to the rapidly growing area of computational psychiatry, as it provides a solid model to detect depression. The results are highly promising in improving the personalized management and early detection of depression thus contributing to the improvement in mental health care practices.
  • Intelligent Farming System: IoT and Adaptive Machine Learning for Sustainable Agriculture
    Kunal Devidas Gaikwad, Sankara Narayanan S, Alagumuthukrishnan. S
    Proceedings of the 2025 11th International Conference on Communication and Signal Processing Iccsp 2025, 2025
    The combination of the Internet of Things (IoT) and Adaptive Machine Learning (AML) presents significant opportunities for precision agriculture by allowing real-time observation, flexible decision-making, and forecasting analyses. This study explores the development of an adaptive ML model for smart agriculture systems, focusing on efficient pest detection and predictive yield analysis. The suggested models intend to enhance resource efficiency and minimize environmental effects by incorporating diverse data sources such as IoT-enabled sensors, satellite images, and past agricultural records. The research highlights the design of scalable, interpretable, and cost-effective ML algorithms tailored for diverse agricultural landscapes. These systems address critical challenges such as real-time data processing, pesticide optimization, and ethical concerns related to data privacy. The findings contribute to advancing sustainable and intelligent farming practices, providing a foundation for future innovations in precision agriculture.
  • From Sensors to Solutions: A Survey on IoT and Machine Learning in Modern Agriculture
    Kunal Devidas Gaikwad, Sankara Narayanan S, Alagumuthukrishnan. S
    2025 International Conference on Data Science Agents and Artificial Intelligence Icdsaai 2025, 2025
    In recent years, the convergence of Internet of Things (IoT) devices and Machine Learning (ML) algorithms has paved the way for precision agriculture. By leveraging IoT sensors and real-time data analysis through ML algorithms, farming practices can be optimized for water usage, soil health, crop prediction, and pest control. This survey explores various IoT devices used in agriculture, the integration of ML models, and presents a comparative analysis for different scenarios. Finally, we provide recommendations for the most optimal configurations in various agricultural settings.
  • Enhanced Diabetic Retinopathyanalysis: Unet-Based Lesion Segmentation Coupled with Mobilenetv2 for Feature Extraction and Efficientnetb0 for Classification
    Sankara Narayanan S, Sayan Kumar Bag, Shivansh Singh
    2024 IEEE International Conference on Intelligent Techniques in Control Optimization and Signal Processing Incos 2024 Proceedings, 2024
    In this research, we address the critical issue of precise Diabetic Retinopathy (DR) diagnosis, a condition of ten leading severe vision impairment or blindness in diabetic patients. Leveraging advanced deep learning models and innovative image processing techniques, our study focuses on accurate retinal image segmentation using a UNet model. This segmentation method delineates lesions and retinal structures effectively. Subsequently, Gabor filters are applied for intricate texture pattern extraction, indicative of diverse retinopathy stages. Integrating MobileNetV2 for feature extraction and EfficientNetB0 for multi-class classification significantly enhances the diagnostic accuracy. Our developed system exhibits a promising 91.2% test accuracy, showcasing its potential in DR diagnosis. While challenges related to varying severity levels persist, our robust framework lays the groundwork for future refinements. By amalgamating sophisticated image segmentation, feature extraction, and classification techniques, our system provides a solid foundation for accurate and timely DR assessment. With continuous enhancements, including the incorporation of more extensive and diverse datasets, our approach holds the promise to revolutionize DR diagnostics. The integration of cutting-edge technology into medical practices underscores the transformative impact of artificial intelligence in the realm of ophthalmology, promising improved patient outcomes.
  • Deep Fake Image Classification Engine Using Inception-ResNet-V1 Network
    Kothandaraman D, S. Sankara Narayanan, Mohamed Iqbal M, Ayesha Yekopalli, Sri Krishnadevarayalu S
    Iccds 2024 International Conference on Computing and Data Science, 2024
    The goal of this project is to develop a real or fake facial image classification system using deep learning techniques. The main focus is on the InceptionResnetVl model pretrained on the VGGFace2 dataset and Kaggle DeepFake Classification Dataset for face classification. The application uses the facenet_pytorch library to perform face detection and preprocessing. Main features of the project include predicting the authenticity of a given facial image as either “real” or “fake”. This is achieved by implementing the InceptionResnetVl model, which is fine-tuned for binary classification with a single output representing the probability of authenticity. The system is designed to run on the GPU when available, enabling faster computing. Grad-CAM (gradient-weighted class activation mapping) technique is used to ensure the explainability of classification decisions. This method creates class activation maps that visually highlight the regions of the input image that influenced the classification decision. The Grad-CAM printout is then superimposed on the original facial image, creating an interpretable visualization of the model's decision-making process. The Gradio user interface is used for the interactive presentation of the project. Users can upload their face and get real vs. false predictions and visual explanations generated by the Grad-CAM algorithm. The user interface also displays a confidence score for each forecast, giving users an idea of how reliable the model is. In general, this project provides a comprehensive system to classify real and fake faces, and the proposed method reaches the accuracy of 97% both in training and testing.
  • Dynamic distributed generative adversarial network for intrusion detection system over internet of things
    S. Balaji, S. Sankara Narayanan
    Wireless Networks, 2023
  • Security and Privacy in Wireless Sensor Networks Using Intrusion Detection Models to Detect DDOS and Drdos Attacks: A Survey
    Kosaraju Chaitanya, Sankara Narayanan
    2023 IEEE International Students Conference on Electrical Electronics and Computer Science Sceecs 2023, 2023
    Wireless sensor networks (WSN) playa significant role in the collection and transmission of data. The principal data collectors and broadcasters are small wireless sensor nodes. As a result of their disorganized layout, the nodes in this network are vulnerable to intrusion. Every aspect of human life includes some form of technological interaction. While the Covid-19 pandemic has been ongoing, the whole corporate and academic world has gone digital. As a direct result of digitization, there has been a rise in the frequency with which Internet-based systems are attacked and breached. The Distributed Denial of Service (DDoS) and Distributed Reflective Denial of Service (DRDoS) assaults are new and dangerous type of cyberattacks that can quickly bring down any service or application that relies on the Internet's infrastructure. Cybercriminals are always refining their methods of attack and evading detection by using techniques that are out of date. Traditional detection systems are not suited to identify novel DDoS attacks since the volume of data created and stored has expanded exponentially in recent years. This research provides a comprehensive overview of the relevant literature, focusing on deep learning for DDoS and DRDoS detection. Due to the expanding number of loT gadgets, distributed DDoS and DRDoS attacks are becoming more likely and more damaging. Due to their lack of generalizability, current attack detection methods cannot be used for early detection of DDoS and DRDoS, resulting in significant load or service degradation when implemented at the endpoint. In this research, a brief review is performed on the models that are used for identification of DDoS and DRDoS attacks. The working of the existing models and the limitations of the models are briefly analyzed in this research.
  • Detection of Breast Cancer using Curvelet Transform and Adaptive Particle Swarm Optimization Technique
    L. C Meena, P. M Joe Prathap, S Sankara Narayanan
    12th IEEE International Conference on Advanced Computing Icoac 2023, 2023
    The breast cancer is the most prevalent malignancy. Better chance of curing breast cancer is early detection, which can also lower mortality rates. The best technique for early breast disease detection is the mammography. In the suggested approach, curvelet transform is utilized to extract features, and adaptive particle swarm optimization helps to choose the eminent features. Adaptive Particle Swarm optimization has been devised to speed up and simplify the process of feature selection and Support Vector Machine (SVM) aids in breast cancer classification. We present an Adaptive Particle Swarm optimization (APSO) that outperforms Particle Swarm optimization (PSO) regarding search efficiency. The suggested model is examined using a collection of 332 images from the Mammographic Image Analysis Society (MIAS) database. The executed findings are compared with the old transforms, and the results demonstrate that the suggested model has higher detection accuracy rates than the earlier approaches.
  • A Reduced Feature-Set OCR System to Recognize Handwritten Tamil Characters using SURF Local Descriptor
    Ashlin Deepa R N, S. Sankara Narayanan, Adithya Padthe, Manjula Ramannavar
    International Journal of Advanced Computer Science and Applications, 2023
    —High dimensionality in variable-length feature sets of real datasets negatively impacts the classification accuracy of traditional classifiers. Convolutional Neural Networks (CNNs) with convolution filters have been widely used for handling the classification of high-dimensional image datasets. However, these models require massive amounts of high-dimensional training data, posing a challenge for many image-processing applications. In contrast, traditional feature detectors and descriptors, with a minor trade-off in precision, have shown success in various computer vision tasks. This paper introduces the Nearest Angles (NA) classifier tailored for a handwritten character recognition system, employing Speeded-Up Robust Features (SURF) as local descriptors. These descriptors make local decisions, while global decisions on the test image are accomplished through a ranking-based classification approach. Image similarity scores generated from the SURF descriptors are ranked to make local decisions, and these ranks are then used by the NA classifier to produce a global class similarity score. The proposed method achieves recognition rates of 96.4% for Tamil, 96.5% for Devanagari, and 97 % for Telugu handwritten character datasets. Although the proposed approach shows slightly lower accuracy compared to CNN-based models, it significantly reduces the computational complexity and the number of parameters required for the classification tasks. As a result, the proposed method offers a computationally efficient alternative to deep learning models, lowering the computational time multiple times without a substantial loss in accuracy
  • Enhancing Glioma Brain Tumor Detection from MRI using Deep Learning Techniques
    S Sankara Narayanan, L C Meena, K Chidambara Thanu, P Chandrasekar
    2023 International Conference on Data Science Agents and Artificial Intelligence Icdsaai 2023, 2023
  • A Neighbor Trust Weight Based Cryptography for Multi Key Distribution for Improving Quality of Service in MANETS
    Rajesh Yamparala, Sankara Narayanan Selvaraj Pandian
    Ingenierie Des Systemes D Information, 2022
  • Hybrid distributed deep-GAN Intrusion detection System in IoT with Autoencoder
    Balaji S., Sankaranarayanan S.
    International Journal of Fuzzy System Applications, 2022
  • Hybrid Deep-Generative Adversarial Network Based Intrusion Detection Model for Internet of Things Using Binary Particle Swarm Optimization
    Balaji S, Dr. S. Sankaranarayanan
    International Journal of Electrical and Electronics Research, 2022
  • Hybrid Deep-GAN Model for Intrusion Detection in IoT Through Enhanced Whale Optimization
    S. Balaji, S. Sankara Narayanan
    International Journal of Computing, 2022
  • Deep Crowd Analysis to Spot Social Distancing Violations in Post-COVID 19 Lifestyle
    S. Sankara Narayanan, A. Nesarani, R. S. Raghav, A. Robert Singh, Suganya Athisayamani
    7th IEEE International Conference on Recent Advances and Innovations in Engineering Icraie 2022 Proceedings, 2022
  • Modified secure AODV protocol to prevent wormhole attack in MANET
    S. Sankara Narayanan, G. Murugaboopathi
    Concurrency and Computation Practice and Experience, 2020
  • Prevention of rushing attack in MANET using threshold-based approach
    S. Sankara Narayanan, G. Murugaboopathi
    International Journal of Internet Technology and Secured Transactions, 2020
  • PFR based technique to detect intruder in MANET
    Journal of Advanced Research in Dynamical and Control Systems, 2020
  • Secure intrusion detection system in mobile Ad hoc networks using RSA algorithm
    S. Sankaranarayanan, G. Murugaboopathi
    Proceedings 2017 2nd International Conference on Recent Trends and Challenges in Computational Models Icrtccm 2017, 2017
  • Secure AODV to combat black hole attack in MANET
    S. Sankara Narayanan, S. Radhakrishnan
    2013 International Conference on Recent Trends in Information Technology Icrtit 2013, 2013