GOPI S

@veltech.edu.in

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

GOPI S


            Download Compact PDF  
https://researchid.co/gopi.scse

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science, Computer Engineering
11

Scopus Publications

Scopus Publications

  • Heart Disease Prediction Using Adaptive Neuro-Fuzzy Inference System (ANFIS) Model with Feature Extraction and Data Pre-Processing
    S. Gopi, M. Arun
    Journal of Multiscale Modelling, 2026
    The new approach came from the research, and its main goal was to create a more accurate model for predicting heart diseases, which would utilize the very powerful combination of data preprocessing, feature extraction, and the latest Adaptive Neuro-Fuzzy Inference System (ANFIS). The application of Principal Component Analysis (PCA) and Mutual Information (MI) not only helped in the reduction of dimensionality and the advancement of feature relevance but also gave support to the model’s learning power. The accuracy achieved through the method proposed was 95.5%, which was supported by 96.0% recall and 94.8% precision as opposed to traditional machine learning and ensemble approaches. The blending of neuro-fuzzy methods contributes to the improvement of diagnosis performance, user-friendliness, and reliability against competing systems, thereby making such a system applicable in diagnosing heart diseases in the medical field.
  • Secure Convergence Model for Distributed Interactions
    Balapu Vishnu Vardhan, Pindiboyana Reethika, Sankar Ganesh Karuppasamy, Sajeev Ram Arumugam, Senthil Murugan V, Gopi S
    7th International Conference on Mobile Computing and Sustainable Informatics Icmcsi 2026, 2026
    In modern digital ecosystems, distributed architectures such as cloud computing, multi-agent systems, and the Internet of Things (IoT) are facing challenges in ensuring a secure and reliable coordination of heterogeneous nodes in the modern digital ecosystem. The current solutions typically cover only one aspect of security, e.g., encryption, trust management, or anomaly detection, without being able to cover the whole spectrum of dynamic networks. This study introduces SCMDI to overcome such drawbacks and it is a secure convergence model for distributed interactions. SCMDI combines hybrid encryption, situational risk assessment, adaptive anomaly detection and policy based access control. It compares the reliability of the nodes with respect to past data and the actual behavior through machine learning models that observe network activity in real time. The framework guarantees confidentiality and integrity by using a combination of symmetric and asymmetric cryptography as well as automatically changing access policies based on the evolving network situations. As well as it is a low weight audit logging system offers tamper free traceability without adding excessive weight to computation. Simulation environment experiments with distributed environments show that SCMDI is effective in terms of detecting malicious activities, imposing trust ratings, and enforcing security policies. In general, the proposed model can provide a high performance, high scalability and intelligent yet secure and collaborative protection to distributed computing environments.
  • Harnessing Quantum Attention: A Hybrid Deep Network for Wearable Sensor-based Activity Recognition
    Purushotham Endla, Asha S, Sravanthi Sallaram, Jayendra Gopal Thatipudi, V.Tejasri, S.Gopi
    Proceedings of 8th International Conference on Computing Methodologies and Communication Iccmc 2025, 2025
    Wearable healthcare sensors enabled Human Activity Recognition (HAR) to become foundational for real-time health monitoring together with individualized medical treatments. The proposed work brings forward QHAR-Net as an application of Quantum Hybrid Attention-Based Deep Network that boosts both recognition precision and operational speed. QHAR-Net achieves multi-modal sensor data pattern recognition through integrating the quantum feature encoding method with ResNet for spatial extraction along with LSTM for temporal dependencies and an attention mechanism. A complete evaluation of the model was performed using a Kaggle database containing thirty selections of ten different human activities. The research demonstrated QHAR-Net reaching 96.5% accuracy which surpassed both CNN-LSTM using 92.1% accuracy and ResNet-LSTM using 94.3% accuracy. Quantum and attention components demonstrated major roles in the overall model performance based on the ablation test. The real-time device testing produced quick inferences within 45 milliseconds using low system resources which indicates suitability for healthcare applications operating from mobile locations. The good results from QHAR-Net research demonstrate its potential to transform portable healthcare monitoring solutions by enhancing measurement precision and processing speed.
  • RetinaVisionMapper: A Framework for Accurate and Early Retinal Disease Identification
    B Prasad, K Lakshmi Prabha, D Shanthi Chelliah, S. Gopi, V. Tejasri, P. Poonkuzhali
    Proceedings of 5th International Conference on Soft Computing for Security Applications Icscsa 2025, 2025
    Retinal diseases such as Diabetic Retinopathy (DR), Glaucoma, and Age-Related Macular Degeneration (AMD) pose severe threats to vision if not identified at early stages. This study presents RetinaVisionMapper, a deep learning based framework which integrates retinal fundus images with physical metadata for precise and early disease detection. The framework relies on a dual-branch designed with a DenseNet-121-based CNN module for fundus imaging and another GRU/MLP stream to handle clinical data. Attention based feature fusion is used to fuse both modalities thus improving cross-modal learning. Advanced preprocessing methods such as CLAHE, non-local means denoising, and vessel enhancement filters are used in order to increase lesion visibility while preserving delicate vascular architecture. Integration of domain-specific data augmentation and Grad-CAM-based interpretability tools is used to achieve robustness and generalizability. The model outperforms 10 baseline algorithms with accuracy 98.72%; precision 98.10%; recall 97.80%; F1-score 97.90% and AUC-ROC 98.95%. These findings suggest that RetinaVisionMapper has therapeutic potential as a clinically viable multi-disease retinal screening tool. The proposed framework provides interpretability, scalability, and precision, suitable for integration into the diagnostic workflow of ophthalmic procedures and in a telemedicine platform for mass screening and early intervention.
  • 5G Approach: Perspectives Into Smart Infrastructure Implementation
    Urinbaev Sharofiddin, S. Gopi, Muntather Almusawi, Rakesh Kumar, Manjinder Kaur Wratch, Amarjeet Kumar Ghosh
    Proceeding of 2024 International Conference on Communication Computing and Energy Efficient Technologies I3ceet 2024, 2024
    Intelligent buildings and intelligent asset monitoring on tenants comfort (ICT for enhanced tenant comfort) Nonetheless, present systems are failing the SFM in connecting so many IoT devices as they tend to have poor connection quality and long latency. 5G technology offers ultra-reliable and low latency, high-speed network infrastructure with slicing features for providing real-time economics. It is also expected to make possible a more sustainable future by reducing energy use and directing new applications in ways that meet ever-tougher sustainability targets. In this blog, he shows the benefits and what is possible now and then explains where 5G has not been fully realised just yet on his SFM applications for various use cases. This document also provides an illustrative example of the Singapore 5G deployment plans based on SUSS and a SFM use case development timeline to be undertaken by the 5G Advanced BIM Lab in alignment with Singapore's 5G deployment strategy. In addition to the 5G forest project, we innovate and co-create with a range of industry partners in order to benefit our students in learning new ways of using this technology across fields and sectors (and also as part of their education/training for entry-level professionals; you read more about it here). Their abilities are supported by development. Facilitates the Utilization of 5G Networks This paper sets a benchmark for future moves on SFM methodology and directs training activities aiming to exploit 5G networks in response to growing sustainability demands, compatible with improved enterprise performance.Keywords- Improved Extreme Learning Machine, Prediction, Human Scenarios, Smart Homes, Healthcare Support, Monitoring, Recurrent Formulation, Feedback Mechanisms, Precision, Swiftness.
  • Secure Vision: Enhancing Data Security in Wireless Sensor Networks through Image Processing
    B. Venkata Swamy, S. Gopi, Kattupalli Sudhakar, Pechetti Girish, M. Janaki Rani, Shrikant Upadhyay
    8th International Conference on I Smac Iot in Social Mobile Analytics and Cloud I Smac 2024 Proceedings, 2024
    The secure Wireless Sensor Network (WSN) architecture was designed in such a way that the tradeoffs among efficiency, scalability and security were balanced. It was consisted of several sensor nodes, cluster heads, base station as well as the control center all of them played a vital role in functioning and managing of data. This network security was provided with a layered security approach like the physical, data link, network, transport, and application layers. Data security and integrity are secured through advanced image processing algorithms, and reliable communication protocol. Regular maintenance, constant monitoring and hardening solutions through resilience and fault tolerance mechanisms added a layer of trust in securing this WSN. A comparative analysis was made using different algorithms, such as Intrusion Detection Systems (IDS) and detection accuracy is 98% and proposed model which has shown a better result in the energy consumption, data transmission latency, encryption time, detection precision as well as efficiency to insure an original data integrity.
  • Enhancing Lifetime in Wireless Sensor Networks through Image Processing-Based Clustering with Genetic Algorithm Routing
    M. Ilampari, M. Amina Begum, N Mahesh Babu, S. Gopi, M. Mythreyee, Shanmugavel Deivasigamani
    Proceedings 2024 4th International Conference on Soft Computing for Security Applications Icscsa 2024, 2024
    Wireless Sensor Networks (WSNs) are a very important technology widely used in diverse application domains of environmental monitoring, disaster management, health care, industrial automation etc. These are networks in which spatially distributed sensor nodes monitor physical or environmental conditions and communications data are processed through a base station. Here, we introduced a novel method for enhancing the efficiency and lifetime of WSNs by using the WSN to process images data and genetic programming (GA) to find the best route. A system model was developed to balance a network consisting of homogenous sensor nodes placed randomly in a particular geographical area, representing realistic scenarios. Each sensor node has limited energy resources, hence the operation of the network was partitioned into clustering phase and routing phase to optimal utilization of energy. To improve both the accuracy and efficiency of clustering, the genetic algorithm then identified the optimal path through these clusters using the best combination of pivot nodes for transmission. The proposed method applies image processing techniques to process sensor data, preprocesses sensor data, and then improves the classification process, clustering results, and the genetic algorithm obtains the more energy-efficient paths. Our extensive evaluation results prove substantial optimizations in network performance, energy efficiency and lifetime over known protocols. This method is expected to be effective in such applications as environmental monitoring, disaster management, and other applications that will utilize reliable and sustainable WSNs.
  • A Secured Multiple Party Key Agreement Protocol Design over Cloud Computing Platform by Using Statistical Data Analysis Logic
    Allam Balaram, Sajja Suneel, P.M. Kavitha, Achinta Saikia, S. Gopi, Y Dileep Kumar
    2nd International Conference on Sustainable Computing and Smart Systems Icscss 2024 Proceedings, 2024
    In the field of cloud computing, ensuring secure and efficient key agreement among multiple parties has emerged as a paramount challenge. Traditional key agreement protocols often rely on central authorities or trusted third parties, posing significant security and privacy concerns. To address these challenges, this paper introduces a novel key agreement protocol designed specifically for cloud computing platforms, emphasizing security, efficiency, and resilience without depending on a trusted third party. The proposed protocol innovatively combines Distributed Key Generation (DKG) with a Dynamic Consensus Mechanism, Zero-Knowledge Proof (ZKP) based authentication, and a Multi-Cloud Redundancy approach, offering a comprehensive solution to secure multi-party communication in distributed cloud environments. The DKG protocol facilitates the collaborative generation of a shared secret among participants, significantly enhancing security by eliminating single points of failure. The proposed Dynamic Consensus Mechanism ensures the integrity and finality of key agreement transactions on a blockchain-based ledger, adapting to network conditions and participant trust levels to optimize performance without compromising security. ZKP-based authentication allows participants to verify their identities without revealing sensitive information, preserving privacy and thwarting impersonation attacks. Lastly, the Multi-Cloud Redundancy strategy enhances the protocol's resilience to cloud-specific vulnerabilities and service outages, ensuring high availability and robustness.
  • Unipolar and Bipolar Mathematical Inference of Weight Adjustment Mode of Single Layer Perceptron on and Logic Gate
    M. Shyamala Devi, S. Goni, P. Tasneem, Karthikeyani Vintha, Dubba Sai Kumar
    2023 3rd International Conference on Advances in Electrical Computing Communication and Sustainable Technologies Icaect 2023, 2023
    In order to ensure that electrical terminals only “switch on” after the proper logic process has been applied, logic gates are employed to make judgments. Each logic gate has a name that explains how various inputs will affect the potential outcomes. To perform logical functions on one or more binary sequence of inputs and produce a single digital output, logic gates are utilized. Embedded systems, microcontrollers, electrical and electronics circuit boards, and programmable logic applications generally make use of logic gates. When the sensor receives no information, it generates low Output impedance for logic 0. It has been determined that using logic gates to create an edge avoider robotics that lowers project budget and speeds up the robot. With the growth of technology, the neural network could also be used for implementing the logic gate using Single Layer Perceptron neural network often termed as Linear Threshold Gate. This paper attempt to analyze the mathematical inference on the number of execution steps for the weight adjustment in the activation function of AND Logic gate. This inference carried out by varying the input of Single layer Perceptron Neural Network in the form of unipolar and bipolar input. The activation function weight adjustment had carried out using batch and sequential mode of execution. Mathematical implementation of AND logic gate exhibits that Bipolar sequential and Bipolar batch mode weight adjustment had less number of execution steps reaching 100% accuracy by matching the desired and predicted output when compared to the unipolar. The detailed execution steps are produced herewith for the validation of outcome.
  • Efficient Packet Flow Path Allocation Using Node Proclivity Tracing Algorithm
    R. Aruna, M. Shyamala Devi, S. Vinoth Kumar, S. Umarani, N. S. Kavitha, S. Gopi
    Lecture Notes in Networks and Systems, 2023
  • Plant Disease Classification using CNN-LSTM Techniques
    E.Anna Devi, S. Gopi, U. Padmavathi, Sajeev Ram Arumugam, S.P. Premnath, Divya Muralitharan
    Proceedings 5th International Conference on Smart Systems and Inventive Technology Icssit 2023, 2023