|
|
The progress in remote photoplethysmography (rPPG) is limited by the key issues of existing publicly available datasets, namely, small size, privacy concerns with facial videos, and single-camera setups. To address these limitations, this paper introduces a comprehensive large-scale multi-view video dataset for rPPG and health parameter estimation. Our dataset includes 3600 video recordings from 600 subjects, captured in both resting states and post-physical activity, using three different web and smartphone cameras at various angles. Each recording is synchronized with a 100 Hz PPG signal and includes additional health metrics, such as arterial pressure, temperature, oxygen saturation, respiratory rate, stress level, and blood test results. |
