My Projects

 Fed-ZSL with LLM-Driven Semantic Attributes for Smart Grid IDSs         Nov 2024 - Present

Working on building a Fed-ZSL based IDS that uses varied semantic attributes to improve data privacy

• Leveraged LLMs to generate distinct semantic attributes for separate clients

Shared Embedding Zero-Shot Federated Learning                 Aug 2024 – Dec 2024

Worked on building an intrusion detection system that can detect unseen attacks simultaneously at various devices, keeping data local

• Developing a Zero-Shot Learning based IDS using Federated Learning.

Power grid Zero-Day Attack Classification using BERT-based Zero-Shot Learning          May 2023 – Jan 2024

Worked on building intrusion detection systems to detect attacks in power grid:

• Developed a Zero-Shot Learning algorithm that can detect new attack types that have not occurred in the past i.e. not present in the training dataset.

 Intrusion Detection for Power Grid: A Review             Published Dec 2023

Dasgupta, R., Pramanik, M., Mitra, P., & Chowdhury, D. R. (2023). Intrusion detection for power grid: a review. International Journal of Information Security, 1-13.

• Wrote a comprehensive review of the existing literature that involves detecting intrusions in power grids, as part of research for MS thesis.

Cervical Cancer Detection Using Hybrid Pooling-Based Convolutional Neural Network Approach          Published April 2023

Mishra, L., Dasgupta, R., Chowdhury, Y. S., Nanda, S., & Nanda, S. (2023). Cervical Cancer Detection Using Hybrid Pooling-Based Convolutional Neural Network Approach. Indian Journal of Gynecologic Oncology, 21(2), 37.

• Proposed pap smear image classification using a hybrid pooling approach, a part of the Bachelor’s degree project. 

Analysis of various optimizer on CNN model in the application of pneumonia detection                                Published May 2021

Chowdhury, Y. S., Dasgupta, R., & Nanda, S. (2021, May). Analysis of various optimizer on CNN model in the application of pneumonia detection. In 2021 3rd International Conference on Signal Processing and Communication (ICPSC) (pp. 417-421). IEEE.

• Analyzed the working of different optimizers in neural networks, a part of the Bachelor’s degree project. 

Performance Comparison of Benchmark Activation Function ReLU, Swish and Mish for Facial Mask Detection Using Convolutional Neural Network       Published July 2021 Dasgupta, R., Chowdhury, Y. S., & Nanda, S. (2021). Performance comparison of benchmark activation function relu, swish and mish for facial mask detection using convolutional neural network. In Intelligent Systems: Proceedings of SCIS 2021 (pp. 355-367). Springer Singapore.

• Compared the working of different activation function in neural networks, a part of the Bachelor’s degree project.