Computer Vision and Pattern Recognition, Information Systems
13
Scopus Publications
18
Scholar Citations
2
Scholar h-index
1
Scholar i10-index
Scopus Publications
A Hybrid AES-GCM Based Structural Encryption Approach for IoT and Alphabet-Restricted Communication Manjunath Kamath K, Akanksh J K, Sujal Sunil Badde, Shambhu K Jha, Vijaya Padmanabha International Conference on Innovative Practices in Technology and Management Iciptm 2026, 2026 The including the IoT networks, cloud infrastruct rapid expansion of interconnected digital ecosystems,ures, and Real-time communication platforms have increased Significantly, there is a need for encryption schemes resistant to current and developing cryptanalytic attacks. Though AES-GCM is still most a popular standard because of its authenticated encryption, speed, robustness and binary ciphertext structure and deterministic nature. Behavior can inadvertently show patterns in repetitive or Streams of structured data. Constrained environments that have high For instance, plaintext redundancy: sensor telemetry, low-entropy IoT traffic, or communication channels with alphabet restrictions, Quite often, this issue becomes very significant. In this paper, a new hybrid architecture A proposal for encryption is provided that extends the traditional AES-GCM by adding a lightweight, reversible transformation layer to perform structural Obfuscation and alphabet-only compatibility. Of the many security features included in the framework are a memory hard Scrypt-based KDF, Base-26 ciphertext encoding, adaptive Vigenre-style substitution, block-level permutations, CSPRNG-driven rotations, deterministic noise injection, and HMAC SHA256 authentication. Experimental evaluations demonstrate high entropy levels <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$5.23-5.33$</tex> bits/char, reliable execution times (¡ 600 ms for 10 KB), complete reversibility, and robust defenses for these: statistical, brute-force, correlation; replay, and tampering attacks. The following proposed system has considerably improved confidentiality and structural unpredictability while remaining appropriate for IoT devices, secure messaging systems, cloud storage, and Environments restricted to alphabet-only payload compatibility.
Design and Implementation of a Secure Log Management System with Integrated Intrusion Detection and Cryptographic Mechanisms Manjunath Kamath K, Krithi G Rao, Advith Rai, Hridya Harindran K, Nitin Pandey, Vijaya Padmanabha International Conference on Innovative Practices in Technology and Management Iciptm 2026, 2026 The LogSec IDS project provides a secure means for storing log file information such that it cannot be easily modified, deleted, or viewed by anyone not authorized to access those logs. Log files hold a significant amount of data pertaining to critical activities that occur on the system. Because of this fact, log files are often targeted by attackers as a means for covering up their activities and remaining undetected within the system. The LogSec IDS Project will implement several methods to ensure confidentiality of the log file contents, as well as integrity of the file. Using AES encryption ensures that the log file is encrypted and cannot be decrypted without the correct decryption key, which is delivered to an authorized user via RSA encryption. Additionally, each log entry will be hashed using chained hashes, so that if any entry in the chain is altered, the connection between the entries will be broken, allowing for easy detection of tampering or other forms of interference. Through periodic integrity verification performed automatically by a cron job, as well as an email notification sent to the log file's administrator upon detection of any tamperings, this system establishes a simple yet very effective automated method for maintaining complete security and accountability of the log file and its entries in real time.
Applying Transfer Learning to Cloud-Based Medical Imaging for Accelerated Diagnostics Manjunath Kamath. K, S. Gomathi Meena, Ananda Thipperudra, V. Malathy, Ravi. G, V. Balamurugan 2nd International Conference on Intelligent Algorithms for Computational Intelligence Systems Iacis 2025, 2025 The rapid growth of medical imaging data presents significant challenges in processing and analysis, especially in resource-constrained healthcare environments. Transfer Learning (TL), a machine learning technique that repurposes pre-trained models, offers a promising solution for addressing these challenges. This study proposes a cloud-based medical imaging framework integrating Transfer Learning to accelerate diagnostic processes while maintaining high accuracy. The system focuses on identifying conditions such as cancer, cardiovascular diseases, and neurological disorders using advanced convolutional neural network (CNN) architectures. The framework employs pre-trained models, including ResNet50 and InceptionV3, to extract meaningful features from medical images. These features are fine-tuned for specific diagnostic tasks, leveraging cloud computing to ensure scalability and computational efficiency. Datasets such as CheXpert and Kaggle's “RSNA Pneumonia Detection Challenge” were used for training and evaluation. Metrics such as training time, diagnostic accuracy, and computational overhead were analyzed to assess performance. The proposed system achieved a 40 % reduction in training time compared to traditional deep learning models, reducing training durations from 10 hours to 6 hours on average. Diagnostic accuracy improved by 25 %, with the framework achieving a top accuracy of 94.3 % across multiple imaging modalities. Additionally, computational overhead was reduced by 20 %, enabling the system to process large datasets efficiently without requiring extensive on-premise infrastructure. The integration of Transfer Learning with cloud-based systems offers a transformative approach to medical imaging diagnostics. By reducing training time and improving accuracy, the proposed framework demonstrates its potential to address the growing demand for efficient and reliable diagnostic solutions. These results highlight the feasibility of deploying TL-powered cloud systems in resource-constrained settings, paving the way for scalable, real-time diagnostic applications in modern healthcare. Future research will focus on extending this approach to multi-modal imaging and real-world clinical implementations.
Olympic Games Analysis and Visualization for Medal Prediction Maheswari Raja, P. Sharmila, P. Vijaya, Roshan Fernandes, Anisha P Rodrigues, Manjunath Kamath K 2025 International Conference on Artificial Intelligence and Data Engineering Aide 2025 Proceedings, 2025 The Olympics is an international sporting event with over 200 countries participating in various competitions. Athletes from different countries compete and make their countries proud of their sporting excellence. Despite their huge populations, many of the most populous nations do not win many Olympic medals. The main purpose of this white paper is to use Python to analyze the Olympics dataset, compare the overall performance of each country, and evaluate each country’s contribution to the Olympics. These analyzes provide greater insight into each nation’s performance at the Olympic Games over the years and help athletes quickly analyze their own and their competitors’ performance. This paper uses exploratory data analysis techniques to compare the performance of different countries and their contribution to the Olympic Games. Various dimensions of the Olympics dataset are visualized to provide a country’s status in the Olympics, help underperforming countries produce quality athletes, and improve their performance in the Olympics.
Optimized Dynamic Spiking Graph Neural Network for Frequency Stabilization in Islanded Grids with Electric Vehicle and Photovoltaic Integration Manjunath Kamath K, Madhu B. K, K. B. Venkata Brahma Rao, Amit Barve, Natrayan L, M. SivaramKrishnan Proceedings 2025 2nd International Conference on Electronic Circuits and Signaling Technologies Icecst 2025, 2025 An islanded Smart Grid (SG) employs distributed resources including photovoltaic (PV) arrays, electric vehicles (EVs), and micro turbines (MTs) to damp deviations in grid frequency by restoring balance among generation, storage, and loads aligned in real time. Nonetheless, the inherently variable and delayed nature of PV and EV outputs introduces a highfrequency disturbance that enlarges the Integral Time Absolute Error (ITAE) and diminishes system efficiency. To mitigate these anomalies, this paper introduces a Dynamic Spiking Graph Neural Network (DSGNN) architecture explicitly tailored for frequency control in islanded circumstances where PV and EV resources are heavily leveraged. The primary objective remains the curtailment of ITAE and the enhancement of efficiency by executing synchronized charging, discharging, and output modulation of EVs and PVs, within timelines shorter than the prevailing frequency feedback delay. DSGNN captures temporally precise graphs of anticipated power and load trajectories through spikecoding and Graph Neural Network fusion, thereby informing control commands that judiciously arrest the evolving frequency error. The resultant architecture thus provides tighter transient frequency bands and accelerates convergence toward the nominal value, thereby fulfilling the islanded grid's operational constraints and availability metrics. The proposed architecture is instantiated and empirically tested within a MATLAB environment. Comparative experiments are executed against established benchmarks namely, Particle Swarm Optimization (PSO), Deep Neural Network (DNN), Artificial Neural Network (ANN), Domain Enriched Navigation (DEN), and Reinforcement Learning (RL). The proposed Deep Structured Guidance Neural Network (DSGNN) achieves an Integral of Timeweighted Absolute Error (ITAE) of 0.2662 along with a steady-state efficiency of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 9. 2 \%}$</tex>. This performance corroborates the hypothesis that the DSGNN reduces frequency deviation in an islanded synchronous grid (SG) augmented with electric vehicle (EV) and photovoltaic (PV) resources, thereby enhancing operational resiliency.
Optimizing Energy Management and Load Balancing Through AI-Driven Quantum Approximate Optimization Manjunath Kamath K, Manasa M, Anitha B. S, Mahadevi, Falguni Tlajiya, Sangramjit Chavan 4th IEEE International Conference on Mobile Networks and Wireless Communications Icmnwc 2024, 2024 The integration of Quantum Approximate Optimization Algorithm (QAOA) and Quantum Annealing (QA) offers a promising approach to addressing energy management and load balancing challenges in modern power systems. This study explores the application of these quantum computing techniques, coupled with Artificial Intelligence (AI), to optimize energy distribution across smart grids. The proposed hybrid model leverages the problem-solving capabilities of QAOA for discrete optimization and the real-time adaptability of QA for minimizing energy consumption and operational costs. AI techniques are employed to predict energy demand, balance loads, and ensure efficient resource allocation. Simulations were conducted on a smart grid scenario involving renewable energy sources, varying demand, and distributed energy generation. The results demonstrate that the hybrid model achieved a 25% improvement in load balancing efficiency and a 30% reduction in energy wastage compared to traditional optimization methods. Additionally, the model reduced peak energy demand by 20%, contributing to overall grid stability. The quantum-enhanced approach also decreased computational complexity by 40%, enabling faster decision-making in dynamic environments. These findings highlight the potential of combining quantum algorithms with AI for sustainable and efficient energy management.
Neuromorphic-Driven Agentic AI for Autonomous Decision-Making Systems Manjunath Kamath K, Samata Mehta.S, Akshaya H. L, Shilpashree N, Girish Jadhav, Abhijit Mitra 4th IEEE International Conference on Mobile Networks and Wireless Communications Icmnwc 2024, 2024 Agentic AI represents a paradigm shift in the development of intelligent systems capable of adaptive and proactive interactions in dynamic and complex environments. By integrating reinforcement learning (RL) with cognitive frameworks, Agentic AI goes beyond traditional rule-based and reactive models, enabling autonomous systems to make informed decisions, anticipate future states, and learn from experience. This paper explores the theoretical foundations and practical applications of Agentic AI, highlighting its potential to transform a variety of fields, including robotics, autonomous driving, finance, and healthcare. Through a detailed review of state-of-the-art research, we illustrate how cognitive architectures such as ACT-R and Soar, combined with advanced RL techniques like Deep Q-Networks (DQN) and Proximal Policy Optimization (PPO), contribute to the development of AI agents with human-like reasoning and decision-making capabilities. Experimental results demonstrate that Agentic AI significantly outperforms conventional AI approaches in terms of adaptability, learning efficiency, and decision accuracy. The findings suggest that Agentic AI offers a robust framework for creating intelligent systems capable of complex problem-solving, long-term planning, and proactive behavior, paving the way for the next generation of AI-driven applications.
A Robust Reversible Data Hiding Framework for Video Steganography Applications Manjunath Kamath K, R. Sanjeev Kunte International Journal of Advanced Computer Science and Applications, 2022 —Reversible Data Hiding (RDH) is a special form of data hiding approach for data integrity and confidentiality protection where the secret image bits (SI) are embedded into Cover Media (CM) by altering its intrinsic pixel attributes. However, in RDH the CM along with the secret message is recovered at the end of computing phase. However, despite of its potential use-cases for enhancing the embedding performance, when it comes to security for various network standards, the traditional RDH mechanisms cannot fully comply with the standards for different set of attacks during the bit-stream transmission scenarios. Therefore, the proposed study contributes towards a computational framework of a robust RDH framework for Video Steganography (VS) which is modeled and simulated under various attack effects and the observation outcome are produced in before and after attack situations to justify the improvement over Embedding Capacity (EC) and Peak Signal-to-Noise Ratio (PSNR) performance for both CM and secret message unlike traditional difference expansion-based methods (DE). The outcome of the study shows that the formulated RDH method not only achieves better reversibility at lower cost of computing but also ensures effective PSNR and imperceptibility outcome for both CM and secret image.
An Enhanced Framework for Implementing Data Analytics and AI for Sustainable Business Growth: Pharmaceutical Supply Chain Case Validation RHR Manjunath Kamath K, Yathiraj G R, Nivedita N M, Manu Y M, Bharath B 2026 IEEE 2nd International Conference on Computing for Sustainability and … , 2026 2026.0
A Hybrid AES-GCM Based Structural Encryption Approach for IoT and Alphabet-Restricted Communication VP Manjunath Kamath K, Akanksh J K, Sujal Sunil Badde, Shambhu K Jha A Hybrid AES-GCM Based Structural Encryption Approach for IoT and Alphabet … , 2026 2026.0
Design and Implementation of a Secure Log Management System with Integrated Intrusion Detection and Cryptographic Mechanisms VP Manjunath Kamath K, Krithi G Rao, Advith Rai, Hridya Harindran K, Nitin ... 5th International Conference on Innovative Practices in Technology and … , 2026 2026.0
Optimized Dynamic Spiking Graph Neural Network for Frequency Stabilization in Islanded Grids with Electric Vehicle and Photovoltaic Integration MSK Manjunath Kamath K , Madhu B, K.B Venkata Brahma Rao, Amit Barve, Natrayan L 2nd International Conference on Electronic Circuits and Signalling … , 2025 2025.0
Applying Transfer Learning to Cloud-Based Medical Imaging for Accelerated Diagnostics VB Manjunath Kamath K, S. Gomati Meena, Ananda Thipperudra, V. Malathy, Ravi. G 2025 2nd International Conference on Intelligent Algorithms for … , 2025 2025.0
Olympic Games Analysis and Visualization for Medal Prediction MKK Maheswari Raja, P. Sharmila, P. Vijaya, Roshan Fernandes, Anisha P Rodrigues 2025 IEEE International Conference on Artificial Intelligence and Data … , 2025 2025.0
Optimized Anti-Interference Dynamic Integral Neural Network Approach for Dementia Prediction in Health Care MKK Pradeep Kumar B P, Ravi Kumar J, Shankar B B Knowledge Based Systems, Elsevier 113723 , 2025 2025.0
Neuromorphic-Driven Agentic AI for Autonomous Decision-Making Systems M Kamath K 4th International Conference on Mobile Networks and Wireless Communications … , 2025 2025.0
Optimizing Energy Management and Load Balancing through AI-Driven Quantum Approximate Optimization M Kamath K 4th International Conference on Mobile Networks and Wireless Communications … , 2025 2025.0 Citations: 13
A Robust Reversible Data Hiding Framework for Video Steganography Applications KK Manjunath, RS Kunte International Journal of Advanced Computer Science and Applications 13 (3) , 2022 2022.0 Citations: 2
Framework for reversible data hiding using cost-effective encoding system for video steganography K Manjunath Kamath, RS Kunte International Journal of Electrical and Computer Engineering (IJECE) 10 (5 … , 2020 2020.0 Citations: 3
Framework for Data Hiding Operation Using Motion Vectors for Effective Imperceptibility Performance K Manjunath Kamath, R Sanjeev Kunte International Conference on Computer Networks and Inventive Communication … , 2019 2019.0
Reversible Video Steganography using Histogram Shifting DRSK Manjunath Kamath K INTERNATIONAL JOURNAL OF ENGINEERING RESEARCH & TECHNOLOGY 5 (20), 1-5 , 2018 2018.0
Reversible Video Steganography using Histogram Shifting RS Kunte
MOST CITED SCHOLAR PUBLICATIONS
Optimizing Energy Management and Load Balancing through AI-Driven Quantum Approximate Optimization M Kamath K 4th International Conference on Mobile Networks and Wireless Communications … , 2025 2025.0 Citations: 13
Framework for reversible data hiding using cost-effective encoding system for video steganography K Manjunath Kamath, RS Kunte International Journal of Electrical and Computer Engineering (IJECE) 10 (5 … , 2020 2020.0 Citations: 3
A Robust Reversible Data Hiding Framework for Video Steganography Applications KK Manjunath, RS Kunte International Journal of Advanced Computer Science and Applications 13 (3) , 2022 2022.0 Citations: 2
An Enhanced Framework for Implementing Data Analytics and AI for Sustainable Business Growth: Pharmaceutical Supply Chain Case Validation RHR Manjunath Kamath K, Yathiraj G R, Nivedita N M, Manu Y M, Bharath B 2026 IEEE 2nd International Conference on Computing for Sustainability and … , 2026 2026.0
A Hybrid AES-GCM Based Structural Encryption Approach for IoT and Alphabet-Restricted Communication VP Manjunath Kamath K, Akanksh J K, Sujal Sunil Badde, Shambhu K Jha A Hybrid AES-GCM Based Structural Encryption Approach for IoT and Alphabet … , 2026 2026.0
Design and Implementation of a Secure Log Management System with Integrated Intrusion Detection and Cryptographic Mechanisms VP Manjunath Kamath K, Krithi G Rao, Advith Rai, Hridya Harindran K, Nitin ... 5th International Conference on Innovative Practices in Technology and … , 2026 2026.0
Optimized Dynamic Spiking Graph Neural Network for Frequency Stabilization in Islanded Grids with Electric Vehicle and Photovoltaic Integration MSK Manjunath Kamath K , Madhu B, K.B Venkata Brahma Rao, Amit Barve, Natrayan L 2nd International Conference on Electronic Circuits and Signalling … , 2025 2025.0
Applying Transfer Learning to Cloud-Based Medical Imaging for Accelerated Diagnostics VB Manjunath Kamath K, S. Gomati Meena, Ananda Thipperudra, V. Malathy, Ravi. G 2025 2nd International Conference on Intelligent Algorithms for … , 2025 2025.0
Olympic Games Analysis and Visualization for Medal Prediction MKK Maheswari Raja, P. Sharmila, P. Vijaya, Roshan Fernandes, Anisha P Rodrigues 2025 IEEE International Conference on Artificial Intelligence and Data … , 2025 2025.0
Optimized Anti-Interference Dynamic Integral Neural Network Approach for Dementia Prediction in Health Care MKK Pradeep Kumar B P, Ravi Kumar J, Shankar B B Knowledge Based Systems, Elsevier 113723 , 2025 2025.0
Neuromorphic-Driven Agentic AI for Autonomous Decision-Making Systems M Kamath K 4th International Conference on Mobile Networks and Wireless Communications … , 2025 2025.0
Framework for Data Hiding Operation Using Motion Vectors for Effective Imperceptibility Performance K Manjunath Kamath, R Sanjeev Kunte International Conference on Computer Networks and Inventive Communication … , 2019 2019.0
Reversible Video Steganography using Histogram Shifting DRSK Manjunath Kamath K INTERNATIONAL JOURNAL OF ENGINEERING RESEARCH & TECHNOLOGY 5 (20), 1-5 , 2018 2018.0
Reversible Video Steganography using Histogram Shifting RS Kunte
