TAMILSELVI M

@roeverengg.edu.in

ASSOCIATE PROFESSOR AND COMPUTER SCIENCE AND ENGINEERING ROEVER ENGINEERING COLLEGE
Dr M TAMILSELVI

TAMILSELVI M


               Download Compact PDF  
https://researchid.co/tamilnaveena

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Engineering, Computer Science, Artificial Intelligence, Decision Sciences
11

Scopus Publications

36

Scholar Citations

2

Scholar h-index

2

Scholar i10-index

Scopus Publications

  • Detection and Taxonomy of Aircraft using Synthetic Aperture Radar Imaging
    D. V. Sairam, M Tamilselvi, S Gayathri Devi, R Bhuvaneswari
    2nd International Conference on Integrated Circuits and Communication Systems Icicacs 2024, 2024
    Improvements in aircraft detection are necessary to improve surveillance. This work investigates the exact detection and classification of airplanes using YOLOv8 in conjunction with Synthetic Aperture Radar (SAR) photos. SAR imaging performs well in dimly lit and unfavorable environments. The objective is to use YOLOv8's object detection to identify aircraft kinds and categories in SAR pictures. A variety of SAR datasets are gathered, preprocessed, and trained on the model to allow for a thorough assessment of YOLOv8's performance in differentiating between different types of aircraft. Measures that assess model performance across aircraft classes include mean average precision (mAP), recall, and loss. The results demonstrate the effectiveness of YOLOv8 and show promising results in terms of speed and precision. This work drives progress in aerial surveillance by highlighting the robustness of SAR when combined with cutting edge deep learning for aircraft taxonomy. The achieved 98.5% mean average precision (mAP) attests to exceptional performance in aircraft detection and categorization from SAR images, reinforcing the utility of this integrated approach in bolstering security systems.
  • Safeguarding the Homeless Using ViTransformer in Violence Detection
    R Bhuvaneswari, Sai Lasya M, B Natarajan, Tamilselvi M, K Geetha, M Velammal
    2024 IEEE International Conference on Intelligent Techniques in Control Optimization and Signal Processing Incos 2024 Proceedings, 2024
    In recent years, the issue of homelessness has gained significant attention as societies grapple with finding solutions to alleviate the plight of those without stable shelter. To address this pressing concern, this research utilizes advanced video analysis techniques, specifically Vision Transformer, to detect the presence of individuals in homeless populations. By leveraging state-of-the-art technology, the research aims to not only identify individuals but also discern their actions and activities within the monitored environment. The methodology involves the initial detection of persons using Vi Transformer, an innovative deep learning model renowned for its accuracy in object recognition tasks. Once a person is successfully identified, this research further scrutinizes their activities, with a particular focus on detecting instances of conflict or fighting. This step holds immense potential or the safety and well-being of homeless individuals, as it enables the swift notification of nearby authorities or support services in cases of altercation. Crucially, this research delves into the performance metrics of both person detection and activity detection experiments, providing valuable insights into the efficacy of the Vi Transformer model in real-world applications. By assessing the system's accuracy, speed, and reliability, this study contributes to the development of proactive strategies for assisting homeless populations, emphasizing the importance of technology in addressing social challenges.
  • Automatic Emotion Detection using SVM-based Optimal Kernel Function
    M Priya, B Ranjani, K Rajammal, N Murali, M Tamilselvi, M Kavitha
    Proceedings of the 18th Indiacom 2024 11th International Conference on Computing for Sustainable Global Development Indiacom 2024, 2024
    Facial expression is very essential in interpersonal relationships. Human-machine interaction is a recently emerging field in which the Facial Expression Recognition System (FERS) is a key aspect in determining the emotional state of a person. This paper proposes a unique way for automatically identifying face expressions from images using Principal Component Analysis (PCA). The proposed approach employs Haar features of the image to detect the face component. The PCA approach is then employed to extract the efficient Haar facial features, resulting in the Eigenvectors, which include dimensionality reduction. A SVM classifier with a kernel function is used to recognize common human emotions such as happiness, fear, sadness, etc. The experimental results demonstrate that the technique proposed in this research is more effective at classifying various expressions and has a higher recognition rate than traditional approaches.
  • Deep Learning Techniques for Fault Detection in Industrial Machinery
    Bhawani Sankar Panigrahi, Thiyagarajan T, M. Tamilselvi, S. B G Tilak Babu, Pavithra G, Bazani Shaik
    5th International Conference on Recent Trends in Computer Science and Technology Icrtcst 2024 Proceedings, 2024
    The use of deep learning techniques for the purpose of improving fault detection in industrial machinery. It is of the utmost importance to have defect detection mechanisms that are both reliable and effective, since the complexity of industrial processes continues to increase. In this paper, the implementation of deep learning algorithms is investigated. These algorithms make use of neural networks to understand complex patterns and anomalies that are present in data coming from machinery. There are many different models that are being researched to see whether or not they are effective in detecting defects at early stages, limiting downtime, and eliminating costly interruptions. These models include convolutional neural networks (CNNs) and recurrent neural networks (RNNs). For the purpose of this , the performance of various methodologies is evaluated over a wide range of industrial situations, taking into consideration issues such as the variability of sensor data and noise. The findings demonstrate the promise of deep learning as a significant tool for enhancing defect detection skills, thereby paving the way for industrial equipment systems that are more reliable and resilient.
  • A Novel Approach for Sign Language Video Generation Using Deep Networks
    Sachin Kumar, Deepa B, Kavitha T, Tamilselvi M, Sathiyapriya V, Natarajan B
    2nd IEEE International Conference on Data Science and Network Security Icdsns 2024, 2024
    Sign language enhances the communication capabilities of the deaf-mute community, allowing for a deeper understanding of their needs and emotions. These languages are highly structured and visual, using gestures and various upper body movements such as those of the hands, face, eyes, and gaze. Researchers face numerous challenges in recognizing and translating the diverse variations in sign movements, which requires specialized expertise in computer vision and artificial intelligence. Sign language recognition and translation research has garnered global attention. This research work introduces a novel methodology for generating sign gesture videos from text inputs by integrating various intelligent techniques. The proposed model employs an enhanced generative adversarial network (GAN) to create sign videos from input sentences. Experiments with the proposed VideoGAN model using diverse sign language datasets from multiple countries have demonstrated its effectiveness. The research outcomes highlight its contribution to high-quality video production, with improved evaluation metrics underscoring the model's superior performance.
  • A Novel Approach for Flower Classification using Deep Networks
    A Jeevan Reddy, B Natarajan, A Abinaya, M Tamilselvi, P Murali, K Geetha
    2nd IEEE International Conference on Data Science and Information System Icdsis 2024, 2024
    The introduction of transformer models, which utilize a self-attention mechanism within deep neural networks, represents a notable breakthrough in natural language processing. This advancement has spurred researchers to investigate its applicability in computer vision tasks. Transformer-based models have showcased remarkable performance compared to traditional convolutional and recurrent neural networks across a spectrum of visual tasks, highlighting their prowess in representation learning. This study aims to classify vision transformer models according to their task-specific functionalities and conducts an extensive evaluation to discern their advantages and limitations. Moreover, it proposes effective strategies for integrating transformers into real-world device-based applications. Investigating the use of vision transformers, inspired by the transformer model, on diverse benchmark datasets, this research focuses on their ability to identify broad-based visual characteristics and distant dependencies through self-attentional mechanisms. Through meticulous examination on the Oxford Flowers dataset, this research delves into various pre-training methods, model topologies, and hyperparameters to enhance recognition accuracy. The findings reveal that vision transformers not only surpass previous state-of-the-art techniques but also exhibit the capacity to discern subtle differences among flower species with remarkable 98.5% accuracy. Furthermore, this research work explores the interpretability of the vision transformer model, elucidating how it recognizes and integrates crucial visual characteristics during classification. By analyzing the model’ s attention maps, insights into its decision-making process are provided. In essence, this research underscores the potential of vision transformers for fine-grained image classification and contributes to the burgeoning field of research on their efficacy in computer vision tasks.
  • A Novel Q-Learning Optimization Approach for Flight Path Prediction in Asian Cities
    Keshavagari Smithin Reddy, B Natarajan, Arthi A, M Tamilselvi, Sridevi R
    2023 3rd Asian Conference on Innovation in Technology Asiancon 2023, 2023
    The domains of logistics and transportation have long been interested in the optimization of flight paths between cities. This research aims to use Skyscanner data to estimate the optimal flight path between 42 Asian destination cities using Reinforcement Learning (RL) techniques, notably Q-learning. RL is a great strategy for addressing the Travelling Salesman Problem (TSP) connected to aircraft route optimization because of the distinctive reward structure it provides. The main objective of the proposed research is to create a model that learns to suggest aircraft routes based on factors such as cost, time, and number of intermediate points that maximize benefits. The proposed research work incorporates a novel Q-learning approach for training an RL agent to predict optimal flight paths. The proposed research work showcases the power-fullness of Q-Learning based RL agents in suggesting optimal flight routes and lays the groundwork for future developments in this research domain. The proposed algorithm provides useful information for tourists and business people looking for accurate and affordable flight path forecasts in the Asian region. The various performance metrics such as Reward Accumulation(RA), Episode Length(EL), and Exploration and Exploitation evaluates the proposed model performance and yields an optimal solution.
  • An Efficient Approach for Obstacle Avoidance and Navigation in Robots
    M Phani Shanmukh, B Natarajan, C. Kannan, M. Tamilselvi, T. Vigneshwaran, S. Syed Husain
    International Conference on Integrated Intelligence and Communication Systems Iciics 2023, 2023
    Reinforcement learning has emerged as a prominent technique for enhancing robot obstacle avoidance capabilities in recent years. This research provides a comprehensive overview of reinforcement learning methods, focusing on Bayesian, static, dynamic policy, Deep Q-Learning (DQN) and extended dynamic policy algorithms. In the context of robot obstacle avoidance, these algorithms enable an agent to interact with its physical environment, learns effective operating strategies, and optimize actions to maximize a reward signal. The environment typically consists of a physical space that the robot must navigate without encountering obstacles. The reward signal serves as an objective measure of the robot’s performance towards accomplishing specific goals, such as reaching designated positions or completing tasks. Furthermore, successful obstacle avoidance strategies acquired in simulation environments can be seamlessly transferred to real-world scenarios. The promising results achieved thus far indicate the potential of reinforcement learning as a powerful tool for enhancing robot obstacle avoidance. This research concludes with insights into the future prospects of reward learning, high-lighting its ongoing importance in the development of intelligent robotics systems. The proposed algorithm DQN outperforms well among all the other algorithms with an accuracy of 81%, Through this research, we aim to provide valuable insights and directions for further advancements in the field of robot obstacle avoidance using reinforcement learning techniques.
  • Enhancing Industrial Load Scheduling through Demand Side Management and Advanced Optimization Techniques Using Neuro-Fuzzy and Deep Learning Approaches
    Shivani Agarwal, Ritu Sharma, M. Tamilselvi, Harsh Mohan Sharma, Dinesh Prasad Sahu, Jagendra Singh
    Proceedings 4th IEEE 2023 International Conference on Computing Communication and Intelligent Systems Icccis 2023, 2023
    This study aims to investigate whether physical factors like temperature, humidity, and pressure interact with current in Micro-electromechanical systems (MEMS). To study these interactions, a straightforward system with four spin-operated engines was employed. The pressure, humidity, and temperature may all change as a result of changing the motor's output current. The results are evaluated and predicted using LSTM and Neuro-Fuzzy models. Sensors are used to keep an eye on these circumstances. Neurofuzzy models employ fuzzy logic and artificial intelligence to capture interconnections, as opposed to LSTM models, which use their capacity to comprehend data connections to capture long-term effects. The results demonstrate that the body's capacity to exert energy is enhanced by the effect of body energy on temperature, humidity, and pressure. The research could lead to improvements in temperature control, energy efficiency, and the functionality and design of massive industrial machinery. A straightforward design is used to launch controlled experiments, and the data collected is then used to train and assess the model. By offering helpful suggestions for the improvement of the motor and the operating electric economy, this study contributes to our understanding of the evolution of the electric motor.
  • Lemuria: A Novel Future Crop Prediction Algorithm Using Data Mining
    M Tamil Selvi, B Jaison
    Computer Journal, 2022
    Agriculture exhibitions an important role in the progression and enlargement of the economy of any country. Prediction of crop yield will be useful for farmers, but it is difficult to predict crop yield because of the climatic factors such as rainfall, soil factors and so on. To tackle these issues, we are implementing a novel algorithm called Lemuria by applying data mining in agriculture especially for crop yield analysis and prediction. This novel algorithm is the hybridization of classifiers for pre-training, training and testing: deep belief network for feature learning, k-means clustering together with particle swarm optimization (PSO) to get the global solution as well as naïve Bayes clustering with PSO for testing. The performance of the Lemuria algorithm is evaluated in Python, which provides an accuracy of 97.74% for crop prediction by considering the rainfall dataset and also stated that this gives the optimum results in comparison with the existing methodologies.
  • Adaptive Lemuria: A progressive future crop prediction algorithm using data mining
    Tamil Selvi M, Jaison B
    Sustainable Computing Informatics and Systems, 2021

RECENT SCHOLAR PUBLICATIONS

  • Design Thinking Based Deep Learning Models For Early and Accurate Detection of HIP Cancer
    NSN Prof. Venkat Namdev Ghodke , Dr Mamta Pathak , Dr.M.Tamilselvi ,Dr. M ...
    IN Patent App. 202,421,002,059 , 2024
    2024
  • Enhancing Industrial Load Scheduling through Demand Side Management and Advanced Optimization Techniques Using Neuro-Fuzzy and Deep Learning Approaches
    JS Shivani Agarwal, Ritu Sharma, M. Tamilselvi, Harsh Mohan Sharma, Dinesh ...
    2023 International Conference on Computing, Communication, and Intelligent … , 2024
    2024
  • An Efficient Approach for Obstacle Avoidance and Navigation in Robots
    SSH M Phani Shanmukh, B Natarajan, C Kannan, M. Tamilselvi, T. Vigneshwaran
    IEEE , 2024
    2024
  • Machine Learning and Data Analytics
    MKS Prof.Yadav Sangeeta Ramachandra,Dr.A.Lokesh,Dr.M.Tamilselvi
    2024
  • AI based Reliable Prediction of Soil Crop Cultivation on Machine Learning Model
    DCG Dr.M.Tamilselvi
    Applications of AI in Emerging Reseach and Education 2, 24-39 , 2023
    2023
  • Interpretation Based on Effectiveness of Blended Mode Learning among Students using Machine Learning Techniques
    MSM Dr.Tamilselvi.M
    Research and Reflection on Education (UGC care approved) 21 (1A), 337-346 , 2023
    2023
  • Interdisciplinary Perspectives on Sustainable Agricultural Management: Integrating Computer Science, Legal Frameworks, and Educational Initiatives”
    DM Tamilselvi
    Tuijin Jishu/Journal of Propulsion Technology 44 (1001-4055), 10 , 2023
    2023
  • Lemuria: a novel future crop prediction algorithm using data mining
    M Tamil Selvi, B Jaison
    The Computer Journal 65 (3), 655-666 , 2022
    2022
    Citations: 16
  • Adaptive Lemuria: A progressive future crop prediction algorithm using data mining
    B Jaison
    Sustainable Computing: Informatics and Systems 31, 100577 , 2021
    2021
    Citations: 20
  • An Overview Of Data Mining Techniques For Maize Yield Prediction
    T M
    IJET 7, 490-92 , 2018
    2018
  • A Malicious And Misbehavior Node Detection Scheme For Wireless Sensor Networks
    T M
    IJRSE 2 (5) , 2018
    2018

MOST CITED SCHOLAR PUBLICATIONS

  • Adaptive Lemuria: A progressive future crop prediction algorithm using data mining
    B Jaison
    Sustainable Computing: Informatics and Systems 31, 100577 , 2021
    2021
    Citations: 20
  • Lemuria: a novel future crop prediction algorithm using data mining
    M Tamil Selvi, B Jaison
    The Computer Journal 65 (3), 655-666 , 2022
    2022
    Citations: 16
  • Design Thinking Based Deep Learning Models For Early and Accurate Detection of HIP Cancer
    NSN Prof. Venkat Namdev Ghodke , Dr Mamta Pathak , Dr.M.Tamilselvi ,Dr. M ...
    IN Patent App. 202,421,002,059 , 2024
    2024
  • Enhancing Industrial Load Scheduling through Demand Side Management and Advanced Optimization Techniques Using Neuro-Fuzzy and Deep Learning Approaches
    JS Shivani Agarwal, Ritu Sharma, M. Tamilselvi, Harsh Mohan Sharma, Dinesh ...
    2023 International Conference on Computing, Communication, and Intelligent … , 2024
    2024
  • An Efficient Approach for Obstacle Avoidance and Navigation in Robots
    SSH M Phani Shanmukh, B Natarajan, C Kannan, M. Tamilselvi, T. Vigneshwaran
    IEEE , 2024
    2024
  • Machine Learning and Data Analytics
    MKS Prof.Yadav Sangeeta Ramachandra,Dr.A.Lokesh,Dr.M.Tamilselvi
    2024
  • AI based Reliable Prediction of Soil Crop Cultivation on Machine Learning Model
    DCG Dr.M.Tamilselvi
    Applications of AI in Emerging Reseach and Education 2, 24-39 , 2023
    2023
  • Interpretation Based on Effectiveness of Blended Mode Learning among Students using Machine Learning Techniques
    MSM Dr.Tamilselvi.M
    Research and Reflection on Education (UGC care approved) 21 (1A), 337-346 , 2023
    2023
  • Interdisciplinary Perspectives on Sustainable Agricultural Management: Integrating Computer Science, Legal Frameworks, and Educational Initiatives”
    DM Tamilselvi
    Tuijin Jishu/Journal of Propulsion Technology 44 (1001-4055), 10 , 2023
    2023
  • An Overview Of Data Mining Techniques For Maize Yield Prediction
    T M
    IJET 7, 490-92 , 2018
    2018
  • A Malicious And Misbehavior Node Detection Scheme For Wireless Sensor Networks
    T M
    IJRSE 2 (5) , 2018
    2018