AI-Based Remote Photoplethysmography: Benchmarking on Realistic Datasets

In this study, we investigate the performance of the Contrast-Phys AI model for remote photoplethysmography on datasets recorded in more realistic conditions, introducing a new dataset named CAMVISIM_LAB for this purpose. The videos in CAMVISIM_LAB were recorded in a lightly controlled lab environment, with participants lying 2 m from the camera and allowed to move their heads, speak, and act naturally. The Contrast-Phys model was evaluated on the CAMVISIM_LAB data and the UBFC-RPPG and PURE datasets, as the latter two are among the most popular datasets for the performance assessment of rPPG methods.The initial evaluation of the heart rate estimation on CAMVISIM_LAB was conducted by training a Contrast-Phys model from scratch using k-fold cross-validation, resulting in a mean absolute error (MAE) of 7.7 bpm. This was improved by using a model pre-trained on the UBFC dataset for rPPG estimation, which improved the result to 5.8 bpm. The error was further minimised by fine-tuning the model with the pretrained weights from UBFC and using our dataset for a second training run. This resulted in an improved MAE of 1.0 bpm when fine-tuning and freezing the initial layers of the ContrastPhys model during training. This result outperformed the score achieved when using the same strategy on the PURE dataset.