Research & Scholarship

2026 International Conference on Cognitive Computing and Networking Systems (ICC-CNS)

My Contribution: I developed a self-supervised dual-stream video anomaly detection framework that combines RGB appearance features and optical flow motion representations using lightweight ResNet-18 backbones and a transformer-based temporal encoder. By integrating temporal prediction with contrastive learning, the proposed framework effectively captures long-range spatio-temporal dependencies, achieving competitive performance on benchmark surveillance datasets while maintaining computational efficiency for real-world applications.

2026 1st International Conference on AI, Data Science, Cyber Security and Smart Manufacturing for Sustainable Development (ICADCS)

My Contribution: I designed a blockchain-based decentralized authentication framework that leverages Ethereum smart contracts, Decentralized Identifiers (DIDs), and QR-code-based credential verification to eliminate reliance on centralized identity providers. The proposed system enables secure credential issuance, real-time verification, and on-chain revocation while ensuring tamper resistance, user privacy, and efficient authentication through a mobile-centric Android implementation.

2026 International Conference on Intelligent Systems in Engineering, Secured Systems and Cybersecurity (ICISESSC)

My Contribution: I proposed a novel multimodal human action recognition framework that integrates infrared heatmap sequences with skeleton-based motion representations using an ActiVit Transformer and HPI-GCN-RP network. By designing a learnable late-fusion strategy, the model achieved robust recognition performance under low-light, occlusion, and cross-view conditions, outperforming individual modality baselines and demonstrating strong generalization on the NTU RGB+D 60 benchmark dataset.

2026 IEEE International Conference on Interdisciplinary Approaches in Technology and Management for Social Innovation (IATMSI)

My Contribution: I developed a novel multimodal fusion technique that combines visible and infrared data to enhance the robustness of human action recognition in challenging environments, achieving state-of-the-art results on benchmark datasets.

2025 12th International Conference on Reliability, Infocom Technologies and Optimization (ICRITO)

My Contribution: My primary contribution was to develop and fine-tuned Sanskrit Tesseract language model, from data collection to model training, resulting in a significant enhancement in OCR accuracy for Sanskrit texts.

2024 Asian Conference on Intelligent Technologies (ACOIT)

My Contribution: My primary contribution was the development and training of the custom Tesseract language model, which involved data collection, image pre-processing, and model fine-tuning, leading to a significant improvement in character recognition accuracy.

2023 International Conference on Communication, Security and Artificial Intelligence (ICCSAI)

My Contribution: I designed and implemented the Convolutional Neural Network (CNN) for vehicle detection and classification, and integrated the model with the IoT hardware for real-time processing.