Biswadeep

Biswadeep

he/him

Ph.D. Candidate in ECE

Posts by Collection

portfolio

publications

Optimal Cache Allocation in a Mobility-Based Heterogeneous Network Using Bayesian Games

Published in IEEE Calcutta Conference, 2017, pp. 423-427, 2017

Introduces a game-theoretic approach to optimally allocate cache in heterogeneous networks, enhancing efficiency and user experience.

Recommended citation: B. Chakraborty, B. Banerjee, A. Mukherjee, M.K. Naskar. (2017). "Optimal Cache Allocation in a Mobility-Based Heterogeneous Network Using Bayesian Games." IEEE Calcutta Conference, 2017, pp. 423-427. https://ieeexplore.ieee.org/document/8280684

Phase-Sensitive Common Spatial Pattern for EEG Classification

Published in IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2019, pp. 3654-3659, 2019

Introduces a phase-sensitive adaptation of the common spatial pattern algorithm for improved EEG signal classification.

Recommended citation: B. Chakraborty, S. Ghosal, L. Ghosh, A. Konar, A.K. Nagar. (2019). "Phase-Sensitive Common Spatial Pattern for EEG Classification." IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2019, pp. 3654-3659. https://ieeexplore.ieee.org/document/8914215

Designing Phase-Sensitive Common Spatial Pattern Filter to Improve Brain-Computer Interfacing

Published in IEEE Transactions on Biomedical Engineering, 2019, Vol. 67, No. 7, pp. 2064-2072, 2019

Introduces a phase-sensitive filter for enhancing the efficacy of brain-computer interfaces, based on the common spatial pattern.

Recommended citation: B. Chakraborty, L. Ghosh, A. Konar. (2019). "Designing Phase-Sensitive Common Spatial Pattern Filter to Improve Brain-Computer Interfacing." IEEE Transactions on Biomedical Engineering, 2019, Vol. 67, No. 7, pp. 2064-2072. https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=10

Phase-Synchrony And Causality Analysis Of Brain Signals To Determine Signal Transduction Pathways In Color Perception

Published in IEEE Region 10 Symposium, 2019, pp. 778-783, 2019

Details an innovative approach for understanding color perception through phase-synchrony and causality analysis of brain signals, identifying key signal transduction pathways.

Recommended citation: S. Ghosal, B. Chakraborty, M. Laha, A. Konar. (2019). "Phase-Synchrony And Causality Analysis Of Brain Signals To Determine Signal Transduction Pathways In Color Perception." IEEE Region 10 Symposium, 2019, pp. 778-783. https://ieeexplore.ieee.org/document/8971255

Design and Evaluation of Self-Assembled Actin-Based Nano-Communication

Published in International Wireless Communications & Mobile Computing Conference (IWCMC), 2019, pp. 208-213, 2019

Explores the feasibility of using actin filaments in designing a novel nano-communication system, focusing on self-assembly mechanisms.

Recommended citation: O.A. Dambri, S. Cherkaoui, B. Chakraborty. (2019). "Design and Evaluation of Self-Assembled Actin-Based Nano-Communication." International Wireless Communications & Mobile Computing Conference (IWCMC), 2019, pp. 208-213. https://ieeexplore.ieee.org/document/8766657

Optimal Selection of EEG Electrodes Using Interval Type-2 Fuzzy-Logic-Based Semi-Separating Signaling Game

Published in IEEE Transactions on Cybernetics, 2020, 2020

Proposes an optimal EEG electrode selection method using interval type-2 fuzzy logic and game theory, improving signal analysis.

Recommended citation: B. Chakraborty, L. Ghosh, A. Konar. (2020). "Optimal Selection of EEG Electrodes Using Interval Type-2 Fuzzy-Logic-Based Semi-Separating Signaling Game." IEEE Transactions on Cybernetics, 2020. https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6221036

Cognitive Analysis of Mental States of People According to Ethical Decisions Using Deep Learning Approach

Published in International Joint Conference on Neural Networks (IJCNN), 2020, pp. 1-8, 2020

Utilizes deep learning to analyze the cognitive processes involved in making ethical decisions, providing insights into human mental states.

Recommended citation: D. Dewan, L. Ghosh, B. Chakraborty, A. Chowdhury, A. Konar, A.K. Nagar. (2020). "Cognitive Analysis of Mental States of People According to Ethical Decisions Using Deep Learning Approach." International Joint Conference on Neural Networks (IJCNN), 2020, pp. 1-8. https://ieeexplore.ieee.org/document/9207555

Published in , 1900

Improving Network Throughput by Hardware Realization of a Dynamic Content Caching Scheme for Information-Centric Networking (ICN)

Published in Wireless Personal Communications, 2021, Vol. 116, No. 4, pp.2873-2898, 2021

Details a dynamic content caching scheme realized through hardware for ICN, aimed at improving network throughput.

Recommended citation: A. Ghosh, B. Chakraborty, A. Raha, A. Mukherjee. (2021). "Improving Network Throughput by Hardware Realization of a Dynamic Content Caching Scheme for Information-Centric Networking (ICN)." Wireless Personal Communications, 2021, Vol. 116, No. 4, pp.2873-2898. https://www.springer.com/journal/11277

Reliable Edge Intelligence in Unreliable Environment

Published in Design, Automation & Test in Europe Conference & Exhibition (DATE), 2021, pp. 896-901, 2021

Examines the challenges and solutions for deploying reliable edge computing intelligence in environments with unreliable connectivity and resources.

Recommended citation: M. Lee, X. She, B. Chakraborty, S. Dash, B. Mudassar, S. Mukhopadhyay. (2021). "Reliable Edge Intelligence in Unreliable Environment." Design, Automation & Test in Europe Conference & Exhibition (DATE), 2021, pp. 896-901. https://ieeexplore.ieee.org/document/9474097

Characterization of Generalizability of Spike Timing Dependent Plasticity trained Spiking Neural Networks

Published in Frontiers in Neuroscience, 2021, 2021

This work explores the generalizability of spiking neural networks (SNNs) trained with spike-timing-dependent plasticity (STDP), revealing insights into their potential for diverse applications.

Recommended citation: B. Chakraborty, S. Mukhopadhyay. (2021). "Characterization of Generalizability of Spike Timing Dependent Plasticity trained Spiking Neural Networks." Frontiers in Neuroscience, 2021. https://www.frontiersin.org/articles/10.3389/fnins.2021.695357/full

A Fully Spiking Hybrid Neural Network for Energy-Efficient Object Detection

Published in IEEE Transactions on Image Processing, 2021, 2021

Demonstrates a fully spiking hybrid neural network architecture that achieves high-efficiency in object detection tasks, marking a significant advancement in energy-efficient computing.

Recommended citation: B. Chakraborty, X. She, S. Mukhopadhyay. (2021). "A Fully Spiking Hybrid Neural Network for Energy-Efficient Object Detection." IEEE Transactions on Image Processing, 2021. https://ieeexplore.ieee.org/abstract/document/9591302

Heterogeneous Recurrent Spiking Neural Network for Spatio-Temporal Classification

Published in Frontiers in Neuroscience, 2022

This paper presents a heterogeneous recurrent spiking neural network (HRSNN) with unsupervised learning for spatio-temporal classification of video activity recognition tasks. Achieving high accuracy across various datasets, the HRSNN model introduces a novel combination of architecture and learning method heterogeneity, offering performance comparable to supervised SNNs with fewer resources.

Recommended citation: Biswadeep Chakraborty, Saibal Mukhopadhyay. (2022). "Heterogeneous Recurrent Spiking Neural Network for Spatio-Temporal Classification." Frontiers in Neuroscience. https://doi.org/10.3389/fnins.2023.99451

MONETA: A Processing-in-Memory-Based Hardware Platform for the Hybrid Convolutional Spiking Neural Network with Online Learning

Published in Frontiers in Neuroscience, 2022, 2022

Introducing MONETA, a pioneering hardware platform that leverages processing-in-memory technology for enhanced efficiency in hybrid convolutional spiking neural networks, supporting online learning capabilities.

Recommended citation: D. Kim, B. Chakraborty, X. She, E. Lee, B. Kang, S. Mukhopadhyay. (2022). "MONETA: A Processing-in-Memory-Based Hardware Platform for the Hybrid Convolutional Spiking Neural Network with Online Learning." Frontiers in Neuroscience, 2022. https://www.frontiersin.org/articles/10.3389/fnins.2022.775457/full

Heterogeneous Recurrent Spiking Neural Network for Spatio-Temporal Classification

Published in Frontiers in Neuroscience, 2022, 2022

Introduces a novel heterogeneous recurrent spiking neural network (HRSNN) using unsupervised learning for video activity recognition, achieving high accuracy with less computational resources and training data, outperforming current homogeneous models.

Recommended citation: B. Chakraborty, S. Mukhopadhyay. (2022). "Heterogeneous Recurrent Spiking Neural Network for Spatio-Temporal Classification." Frontiers in Neuroscience, 2022. https://www.frontiersin.org/articles/10.3389/fnins.2023.994517/full

Brain-Inspired Spatiotemporal Processing Algorithms for Efficient Event-Based Perception

Published in 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE), 2023

This conference presentation discusses the advancement of neuromorphic event-based cameras through brain-inspired algorithms, particularly spiking neural networks (SNNs), to enable efficient, low-power, and low-latency spatiotemporal processing for real-world applications. Highlighting the utility of SNNs in approximating spatiotemporal sequences and their event-driven training and inference for real-time performance, it also explores the role of associative memory structures in enhancing event-based perception.

Recommended citation: Biswadeep Chakraborty, Uday Kamal, Xueyuan She, Saurabh Dash, Saibal Mukhopadhyay. (2023). "Brain-Inspired Spatiotemporal Processing Algorithms for Efficient Event-Based Perception." Presented at 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE), IEEE, pp. 1-6.

Unsupervised 3D Object Learning through Neuron Activity Aware Plasticity

Published in International Conference on Learning Representations (ICLR), 2023, 2023

Presents an innovative approach for unsupervised learning of 3D objects leveraging neuron activity-aware plasticity mechanisms.

Recommended citation: B. Kang, B. Chakraborty, S. Mukhopadhyay. (2023). "Unsupervised 3D Object Learning through Neuron Activity Aware Plasticity." International Conference on Learning Representations (ICLR), 2023. #

A Reconfigurable Quantum State Tomography Solver in FPGA

Published in IEEE International Conference on Quantum Computing & Engineering (QCE), 2023, 2023

Discusses the development of a flexible FPGA-based solution for quantum state tomography, facilitating enhanced quantum computing applications.

Recommended citation: N. Miller, B. Chakraborty, S. Mukhopadhyay. (2023). "A Reconfigurable Quantum State Tomography Solver in FPGA." IEEE International Conference on Quantum Computing & Engineering (QCE), 2023. #

XMD: An expansive Hardware-telemetry based Mobile Malware Detector for Endpoint Detection

Published in IEEE Transactions on Information Forensics and Security 2023, 2023

Details a novel approach to mobile malware detection using hardware telemetry, enhancing endpoint security.

Recommended citation: H. Kumar, B. Chakraborty, S. Sharma, S. Mukhopadhyay. (2023). "XMD: An expansive Hardware-telemetry based Mobile Malware Detector for Endpoint Detection." IEEE Transactions on Information Forensics and Security 2023. https://ieeexplore.ieee.org/abstract/document/10262063

Brain-Inspired Spiking Neural Network for Online Unsupervised Time Series Prediction

Published in 2023 International Joint Conference on Neural Networks (IJCNN), 2023

Introducing the Continuous Learning-based Unsupervised Recurrent Spiking Neural Network Model (CLURSNN), this paper tackles the challenge of energy and data-efficient online time series prediction in evolving dynamical systems, crucial for edge AI applications. Unlike traditional DNN-based models requiring extensive data and continuous retraining, CLURSNN leverages spike-timing dependent plasticity (STDP) for online predictions, adapting to changes in the underlying system efficiently and without the need for retraining.

Recommended citation: Biswadeep Chakraborty, Saibal Mukhopadhyay. (2023). "Brain-Inspired Spiking Neural Network for Online Unsupervised Time Series Prediction." 2023 International Joint Conference on Neural Networks (IJCNN), IEEE, pp. 1-8.

STEMFold: Stochastic Temporal Manifold for Multi-Agent Interactions in the Presence of Hidden Agents

Published in arXiv preprint arXiv:2401.14522, 2024

STEMFold introduces a spatiotemporal attention-based generative model designed to predict the unmeasured dynamics of multi-agent systems from observations of visible agents. Leveraging a novel spatiotemporal graph with time anchors, it maps observations to a stochastic manifold without prior knowledge of the interaction graph topology, demonstrating superior performance in predicting complex multi-agent interactions in the presence of hidden agents.

Recommended citation: Hemant Kumawat, Biswadeep Chakraborty, Saibal Mukhopadhyay. (2024). "STEMFold: Stochastic Temporal Manifold for Multi-Agent Interactions in the Presence of Hidden Agents." arXiv preprint arXiv:2401.14522. https://arxiv.org/abs/2401.14522

Studying the Impact of Stochasticity on the Evaluation of Deep Neural Networks for Forest-Fire Prediction

Published in arXiv preprint arXiv:2402.15163, 2024

This paper embarks on the first systematic study to evaluate Deep Neural Networks (DNNs) for forest-fire prediction under stochastic conditions. By developing a framework that distinguishes between classification-based metrics and proper scoring rules, the study unveils the reliability of fidelity-to-statistic evaluations in stochastic scenarios. The research further applies this analytical lens to real-world wildfire data, uncovering the shortcomings of conventional evaluation techniques and proposing stochasticity-compatible, interpretable alternatives for wildfire prediction models.

Recommended citation: Harshit Kumar, Biswadeep Chakraborty, Beomseok Kang, Saibal Mukhopadhyay. (2024). "Studying the Impact of Stochasticity on the Evaluation of Deep Neural Networks for Forest-Fire Prediction." arXiv preprint arXiv:2402.15163.

Topological Representations of Heterogeneous Learning Dynamics of Recurrent Spiking Neural Networks

Published in arXiv preprint arXiv:2403.12462, 2024

This paper explores the unique topological representations learned by Recurrent Spiking Neural Networks (RSNNs) through unsupervised learning methods like spike-timing dependent plasticity (STDP). Introducing a novel application of Representation Topology Divergence (RTD) to RSNNs, the study reveals how heterogeneous learning dynamics contribute to the development of distinct neural representations, offering new insights into the capabilities of SNNs and their potential in creating more biologically plausible hybrid AI systems.

Recommended citation: Biswadeep Chakraborty, Saibal Mukhopadhyay. (2024). "Topological Representations of Heterogeneous Learning Dynamics of Recurrent Spiking Neural Networks." arXiv preprint arXiv:2403.12462.

Sparse Spiking Neural Network: Exploiting Heterogeneity in Timescales for Pruning Recurrent SNN

Published in International Conference on Learning Representations, 2024, 2024

A task-agnostic pruning method that exploits the diversity in timescales for heterogeneous RSNNs and gives small, stable pruned networks.

Recommended citation: B. Chakraborty, B. Kang, H. Kumar, S. Mukhopadhyay. (2024). "Sparse Spiking Neural Network: Exploiting Heterogeneity in Timescales for Pruning Recurrent SNN." International Conference on Learning Representations, 2024. https://openreview.net/forum?id=0jsfesDZDq

RT-HMD: A Novel Statistical Approach for Robust Real-Time Hardware-based Malware Detection under Weak Supervision Formulation

Published in Submitted in International Symposium on Low Power Electronics and Design (ISLPED) 2024, 2024

RT-HMD introduces a groundbreaking statistical methodology for real-time, hardware-based malware detection, leveraging weak supervision to achieve robustness and efficiency. This novel approach addresses the challenges of detecting sophisticated malware in real-time with limited supervision, offering a promising solution for enhancing cybersecurity measures in critical hardware systems.

Recommended citation: Harshit Kumar, Sudarshan Sharma, Biswadeep Chakraborty, Saibal Mukhopadhyay. (2024). "RT-HMD: A Novel Statistical Approach for Robust Real-Time Hardware-based Malware Detection under Weak Supervision Formulation." Submitted in ISLPED 2024.

talks

Phase-Sensitive Common Spatial Pattern for EEG Classification

Published:

This talk at the IEEE International Conference on Systems, Man, and Cybernetics (SMC) presented a novel approach to EEG signal classification using a phase-sensitive common spatial pattern. The methodology enhances the ability to differentiate between various mental states, offering improved performance in brain-computer interface applications. The discussion included results from recent studies and potential implications for future research.

Learning to Predict Using Network of Spiking Neurons

Published:

Presented at the ICERM Workshop, this talk explored the novel approaches of using networks of spiking neurons for predictive modeling. It covered the theoretical underpinnings as well as practical applications, demonstrating how spiking neural networks can offer significant advantages in processing speed and efficiency for machine learning tasks.

Dynamics in Diversity: Harnessing Heterogeneity in Neuronal and Synaptic Dynamics in SNNs

Published:

This invited talk and poster presentation at the SNUFA conference delved into the diverse dynamics of neuronal and synaptic activities within Spiking Neural Networks (SNNs). It highlighted the importance of harnessing this heterogeneity for the development of more efficient and capable neural systems, presenting recent findings and methodologies in the field.

Evolution Strategies in Heterogenous Recurrent Spiking Neural Network for Dynamical Control

Published:

The spotlight talk presented at the Georgia Tech DCL Symposium revolved around the application of evolution strategies in heterogeneous recurrent spiking neural networks. The aim was to explore the potentials of these networks in dynamical control systems, emphasizing the integration of heterogeneity in neuronal and synaptic dynamics to enhance learning and adaptation in complex environments.

Brain-Inspired Spiking Neural Networks for Online Learning

Published:

This talk focused on giving an motivation for brain-inspired computing and neuromorphic computing to undergraduate students and discuss the capabilities of Spiking Neural Networks (SNNs) for advancing online learning mechanisms. The discussion included the temporal dynamics and computational advantages of SNNs, aiming to illustrate how these biological-inspired systems can lead to more efficient and robust artificial intelligence technologies.

teaching

Digital Hardware Design for Non-ECE Majors

Undergraduate course, Georgia Institute of Technology, School of Electrical and Computer Engineering, 2024

In Fall 2024, I will have the opportunity to serve as the Instructor of Record for ECE 2020, focusing on Digital Hardware Design, specifically tailored for non-ECE majors. This role involves not just lecturing, but also comprehensive course management including the creation of the syllabus, scheduling, assignment development, and examination.