Bithi Banik
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Bithi Banik has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Debasish Ghose
Kristiania University of Applied Sciences
3 shared publications - Yuan Lin
School of Economics, Innovation, and Technology, Kristiania University College, Oslo, Norway
2 shared publications - Andrii Shalaginov
School of Technology and Innovation, Kristiania University of Applied Sciences, Kirkegata 24, 0153 Oslo, Norway
1 shared publication - Bjarte Nerland
Kristiania University of Applied Sciences, Bergen, Norway
1 shared publication - Carl-Eirik Dahl Johnsen
Kristiania University of Applied Sciences, Bergen, Norway
1 shared publication - Kim Andy Ystebø
Kristiania University of Applied Sciences, Bergen, Norway
1 shared publication
Latest Publications
- Open Access
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
FedBalance: An Adaptive Cross-Silo Federated Learning Approach for Stress Detection under non-IID Data Distributions
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessFedBalance: An Adaptive Cross-Silo Federated Learning Approach for Stress Detection under non-IID Data Distributions
- Open Access
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Adversarially Robust Federated Learning for IoMT-based Physiological Condition Monitoring: A Benchmark Dataset and ML-Aided Aggregation Framework
Federated learning (FL) is increasingly employed in Internet of Medical Things (IoMT) systems to enable collaborative model training while preserving the privacy of sensitive physiological data. Despite its advantages, IoMT-oriented FL remains highly vulnerable to data poisoning, model poisoning, and Byzantine attacks due to device heterogeneity, limited computational resources, and open client participation, challenges that existing robust aggregation techniques do not fully address. In this work, we introduce a benchmark dataset specifically tailored for IoMT environments, enabling systematic evaluation of adversarial behaviors through high-dimensional feature representations derived from client-side model updates. Using this benchmark, we propose a machine-learning (ML)-assisted aggregation framework that incorporates a malicious client detection module based on geometric and statistical properties of model weights, allowing adversarial updates to be identified and filtered prior to global aggregation. Extensive experiments conducted on ten resource-constrained Internet of Things devices, i.e., Raspberry Pi platforms, demonstrate that the proposed detection module achieves up to 99% accuracy in multi-class attack identification. Furthermore, the proposed ML-assisted aggregation framework consistently outperforms existing robust aggregation methods, exhibiting faster convergence and sustained model accuracy across communication rounds under adversarial conditions.
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessAdversarially Robust Federated Learning for IoMT-based Physiological Condition Monitoring: A Benchmark Dataset and ML-Aided Aggregation Framework
Federated learning (FL) is increasingly employed in Internet of Medical Things (IoMT) systems to enable collaborative model training while preserving the privacy of sensitive physiological data. Despite its advantages, IoMT-oriented FL remains highly vulnerable to data poisoning, model poisoning, and Byzantine attacks due to device heterogeneity, limited computational resources, and open client participation, challenges that existing robust aggregation techniques do not fully address. In this work, we introduce a benchmark dataset specifically tailored for IoMT environments, enabling systematic evaluation of adversarial behaviors through high-dimensional feature representations derived from client-side model updates. Using this benchmark, we propose a machine-learning (ML)-assisted aggregation framework that incorporates a malicious client detection module based on geometric and statistical properties of model weights, allowing adversarial updates to be identified and filtered prior to global aggregation. Extensive experiments conducted on ten resource-constrained Internet of Things devices, i.e., Raspberry Pi platforms, demonstrate that the proposed detection module achieves up to 99% accuracy in multi-class attack identification. Furthermore, the proposed ML-assisted aggregation framework consistently outperforms existing robust aggregation methods, exhibiting faster convergence and sustained model accuracy across communication rounds under adversarial conditions.
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- p. 1
FedSmartCare: Design and Implementation of a Federated Learning Enabled Vital Sign Monitoring System
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- p. 1
Open AccessFedSmartCare: Design and Implementation of a Federated Learning Enabled Vital Sign Monitoring System