Population
Facial videos from the PURE and UBFC-rPPG datasets
Comparison
STC-rPPG network vs existing state-of-the-art models
Key result
The proposed STC-rPPG network achieved high accuracy for remote heart rate estimation, with mean absolute errors of 0.23 bpm and 0.49 bpm on the PURE and UBFC-rPPG datasets, respectively.
Authors
Loading...
STC-rPPG may improve remote heart rate estimation; leaves open clinical validation before practice adoption.
The proposed STC-rPPG network improves remote heart rate estimation accuracy from facial videos by effectively modeling spatiotemporal dynamics and inter-channel correlations.
Lin et al. (2025) studied Heart rate estimation. STC-rPPG network vs. Existing state-of-the-art models was evaluated on Heart rate estimation accuracy (MAE and RMSE). The proposed STC-rPPG network achieved high accuracy for remote heart rate estimation, with mean absolute errors of 0.23 bpm and 0.49 bpm on the PURE and UBFC-rPPG datasets, respectively.