PulseBench: Benchmarking Heart Rate and Heart Rate Variability Estimation from PPG Signals
We will develop a reproducible benchmarking framework for estimating heart rate and heart-rate variability from photoplethysmography signals. The project will evaluate deep learning methods including self-supervised methods across diverse datasets, focusing on accuracy, robustness, and practical deployment constraints. The project combines biomedical signal processing, wearable sensing, deep learning learning, benchmarking, and real-world health-monitoring applications.




