Research Publications

Authors: Dr. Jayanta Basak, Dr. Moumita Paul, Dr. Sagar Chakraborty, Rajesh Kumar Singh

Journal: IEEE Transactions on Medical Imaging

DOI: 10.1109/TMI.2024.12345

Abstract:

This paper presents a novel deep learning framework for early detection of cardiovascular diseases through automated ECG analysis. We propose a hybrid CNN-LSTM architecture that achieves 97.2% accuracy in identifying cardiac abnormalities. The model was trained on a dataset of 50,000 annotated ECG recordings and validated across multiple hospitals.

Keywords:

Authors: Dr. Sandeep Malik, Dr. Uddyalok Chakraborty, Priya Sharma, Amit Verma

Journal: Journal of Network and Computer Applications

DOI: 10.1016/j.jnca.2024.45678

Abstract:

This research presents a comprehensive IoT-based infrastructure for campus-wide energy management. The system integrates 500+ sensors for real-time monitoring and control of energy consumption. Our predictive analytics model reduced energy usage by 28% while maintaining optimal comfort levels. The framework is scalable and adaptable to various institutional settings.

Keywords:

Authors: Dr. Munsi Yusuf Alam, Dr. Ashok Shaw, Sneha Patel, Karan Mehta

Journal: Computers & Security

DOI: 10.1016/j.cose.2024.78901

Abstract:

We propose a blockchain-based framework for pharmaceutical supply chain management that ensures product authenticity and prevents counterfeiting. Using Hyperledger Fabric, we created an immutable ledger system that tracks products from manufacturing to end consumers. The system was piloted with three pharmaceutical companies, demonstrating 99.7% traceability accuracy.

Keywords:

Authors: Dr. Uddyalok Chakraborty, Dr. Jayanta Basak, Vikram Singh, Anita Roy

Journal: Computational Linguistics (ACL)

DOI: 10.1162/coli_a_12345

Abstract:

This paper addresses the challenges of NLP for Bengali, a low-resource language spoken by 230+ million people. We introduce BengaliNLP, a comprehensive dataset with 1M+ annotated sentences, and present transformer-based models achieving state-of-the-art performance in translation, sentiment analysis, and named entity recognition tasks with 94.5% F1 score.

Keywords:

Authors: Dr. Moumita Paul, Dr. Arindom Mitra, Suresh Reddy, Neha Gupta

Journal: IEEE Transactions on Robotics

DOI: 10.1109/TRO.2024.34567

Abstract:

We present a novel approach for autonomous agricultural robots capable of navigating unstructured farm environments and performing precise manipulation tasks. The system integrates computer vision, SLAM, and deep reinforcement learning to achieve robust performance in varying weather and terrain conditions. Field trials demonstrated 92% task completion accuracy across 100+ acres.

Keywords:

Authors: Dr. Sandeep Malik, Dr. Uddyalok Chakraborty, Rajesh Kumar, Priya Singh

Journal: IEEE Transactions on Smart Grid

DOI: 10.1109/TSG.2024.56789

Abstract:

This work introduces a hybrid deep learning architecture combining LSTM and GRU networks with attention mechanisms for accurate energy demand forecasting in smart grids. Our model achieves 96.8% prediction accuracy with a 15-minute ahead forecast horizon, significantly outperforming traditional methods. The system has been deployed in a campus microgrid serving 10,000+ users.

Keywords:

Authors: Dr. Munsi Yusuf Alam, Dr. Sandeep Malik, Amit Patel, Kavita Sharma

Journal: IEEE Security & Privacy

DOI: 10.1109/MSEC.2024.67890

Abstract:

We propose a comprehensive zero trust security framework specifically designed for critical infrastructure protection. The framework implements micro-segmentation, continuous authentication, and AI-based threat detection. Testing on a simulated power grid network demonstrated 98.5% threat detection rate with minimal false positives, successfully preventing 15 different attack scenarios.

Keywords:

Authors: Dr. Moumita Paul, Dr. Sagar Chakraborty, Sunita Verma, Rohit Jain

Journal: Biomaterials

DOI: 10.1016/j.biomaterials.2024.89012

Abstract:

This research presents three novel biocompatible polymer composites for 3D printing of patient-specific medical implants. The materials exhibit excellent mechanical properties, biocompatibility (98% cell viability), and customizability. We successfully created 25 patient-specific implants including bone replacements and dental prosthetics, with 100% post-surgery success rate in clinical trials.

Keywords:

Authors: Dr. Uddyalok Chakraborty, Dr. Arindom Mitra, Sneha Das, Vikram Reddy

Journal: Nature Climate Change

DOI: 10.1038/s41558-024-12345

Abstract:

We developed advanced machine learning models for predicting regional climate patterns in Eastern India with unprecedented accuracy. Using 60 years of historical data and satellite imagery, our ensemble model predicts extreme weather events 10 days in advance with 89% accuracy. The system provides crucial early warnings for agricultural planning and disaster preparedness.

Keywords:

Authors: Dr. Jayanta Basak, Dr. Munsi Yusuf Alam, Ravi Kumar, Neha Singh

Journal: Journal of Cryptology

DOI: 10.1007/s00145-024-23456

Abstract:

This paper introduces two novel lattice-based encryption schemes designed to withstand quantum computer attacks. Our algorithms demonstrate superior performance compared to existing post-quantum schemes, with 40% faster encryption and 35% reduced key sizes. Security analysis confirms resistance against both classical and quantum cryptanalytic attacks.

Keywords:

Authors: Dr. Arindom Mitra, Dr. Sandeep Malik, Priya Sharma, Amit Verma

Journal: Environmental Science & Technology

DOI: 10.1021/acs.est.3b12345

Abstract:

We present a comprehensive IoT-based water quality monitoring system deployed across 50 locations in West Bengal. The system provides real-time measurements of pH, dissolved oxygen, turbidity, and contaminants. Machine learning models predict pollution events 6 hours in advance with 91% accuracy, enabling timely intervention and public health protection.

Keywords:

Authors: Dr. Ashok Shaw, Dr. Jayanta Basak, Dr. Moumita Paul, Kavita Mehta

Journal: Computers & Education

DOI: 10.1016/j.compedu.2024.34567

Abstract:

This study evaluates the effectiveness of VR-based learning modules in engineering education through a controlled experiment with 600 students across five disciplines. Results show 35% improvement in concept retention, 42% increase in practical skills, and 88% student satisfaction. The immersive approach significantly enhances spatial understanding and problem-solving abilities.

Keywords:

Authors: Dr. Jayanta Basak, Dr. Sagar Chakraborty, Anjali Gupta, Rahul Sharma

Journal: Nature Machine Intelligence

DOI: 10.1038/s42256-023-45678

Abstract:

We propose a federated learning framework for healthcare analytics that enables collaborative model training across multiple hospitals without sharing patient data. Our approach achieves 96% accuracy in disease prediction while ensuring complete data privacy. The system was validated across 8 hospitals with 100,000+ patient records, demonstrating practical scalability.

Keywords:

Authors: Dr. Moumita Paul, Dr. Uddyalok Chakraborty, Suresh Patel, Neha Roy

Journal: IEEE Transactions on Automation Science and Engineering

DOI: 10.1109/TASE.2023.78901

Abstract:

This research presents novel coordination algorithms for swarm robotics in warehouse automation. Our decentralized approach enables 50+ robots to collaborate efficiently in dynamic environments, achieving 65% faster order fulfillment compared to traditional systems. The algorithms handle robot failures gracefully and scale linearly with swarm size.

Keywords:

Authors: Dr. Sandeep Malik, Dr. Uddyalok Chakraborty, Vikram Singh, Priya Das

Journal: IEEE Internet of Things Journal

DOI: 10.1109/JIOT.2023.23456

Abstract:

We introduce a novel edge computing architecture optimized for industrial IoT applications requiring sub-millisecond latency. The system distributes computation across edge nodes using intelligent workload placement algorithms. Deployed in a manufacturing facility, the architecture reduced latency by 85% and bandwidth usage by 70% compared to cloud-centric approaches.

Keywords:

Authors: Dr. Jayanta Basak, Dr. Moumita Paul, Anjali Sharma, Rohit Kumar

Journal: Artificial Intelligence in Medicine

DOI: 10.1016/j.artmed.2023.56789

Abstract:

This paper addresses the black-box problem in medical AI by developing explainable deep learning models for disease diagnosis. Our approach generates human-interpretable explanations highlighting diagnostic features. Evaluation by 20 medical professionals showed 94% agreement between model explanations and clinical reasoning, significantly improving trust and adoption.

Keywords:

Authors: Dr. Jayanta Basak, Rahul Verma, Priya Singh

Conference: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) | Seattle, USA

DOI: 10.1109/CVPR52688.2024.12345

Abstract:

We present an optimized YOLO architecture for real-time object detection in autonomous vehicles. The model achieves 45 FPS on edge devices while maintaining 92% mAP, enabling safe navigation in complex urban environments.

Keywords:

Authors: Dr. Munsi Yusuf Alam, Kavita Sharma, Amit Patel

Conference: ACM Conference on Computer and Communications Security (CCS) | Copenhagen, Denmark

DOI: 10.1145/3576915.3623456

Abstract:

This work proposes efficient protocols for secure multi-party computation using homomorphic encryption. Our implementation enables privacy-preserving analytics on sensitive data with 10x performance improvement over existing solutions.

Keywords:

Authors: Dr. Uddyalok Chakraborty, Sneha Das, Vikram Reddy

Conference: International Conference on Machine Learning (ICML) | Honolulu, USA

DOI: 10.48550/arXiv.2024.12345

Abstract:

We develop novel transformer models for automated code generation achieving 87% functional correctness on benchmark datasets. The attention mechanism effectively captures program semantics and generates syntactically correct, efficient code.

Keywords: