A5101900210- Advanced Neural Network Applications
- IoT and Edge/Fog Computing
- Water Quality Monitoring Technologies
- EEG and Brain-Computer Interfaces
- Advanced Data Storage Technologies
- Underwater Vehicles and Communication Systems
- Semiconductor materials and devices
- Data Stream Mining Techniques
- IoT-based Smart Home Systems
- Ferroelectric and Negative Capacitance Devices
- Neural dynamics and brain function
- Functional Brain Connectivity Studies
University College London
2024
Imperial College London
2023-2024
Dyson (United Kingdom)
2024
Great Ormond Street Hospital
2024
In the field of medical diagnosis and patient monitoring, effective pattern recognition in neurological time-series data is essential. Traditional methods predominantly based on statistical or probabilistic learning inference often struggle with multivariate, multi-source, state-varying, noisy while also posing privacy risks due to excessive information collection modeling. Furthermore, these overlook critical information, such as distribution points inherent uncertainties. To address...
Machine learning (ML) is moving towards edge devices. However, ML models with high computational demands and energy consumption pose challenges for inference in resource-constrained environments, such as the deep sea. To address these challenges, we propose a battery-free model personalization pipeline microcontroller units (MCUs). As an example, performed fish image recognition ocean. We evaluated compared accuracy, runtime, power, of before after optimization. The results demonstrate that,...
The personalization of machine learning (ML) models to address data drift is a significant challenge in the context Internet Things (IoT) applications. Presently, most approaches focus on fine-tuning either full base model or its last few layers adapt new data, while often neglecting energy costs. However, various types exist, and may not result optimal performance certain scenarios. We propose Target Block Fine-Tuning (TBFT), low-energy adaptive framework designed for resource-constrained...