Why the study?
Existing deep learning methods for remote photoplethysmography lack task-specific feature refinement across spatial and channel dimensions, which can dilute pulsatile information.
Population
Facial videos from PURE, UBFC-rPPG, and MMPD datasets
Comparison
MDFS and SCOPE modules vs existing deep learning methods
Key result
The proposed rPPG framework integrating MDFS and SCOPE modules achieved state-of-the-art heart rate estimation on the PURE dataset with a mean absolute error of 0.27 and Pearson correlation of 0.99.
Authors
Loading...
May advance non-contact HR monitoring feasibility; leaves open prospective clinical validation.
Absolute Event Rate: 0.27% vs 0.29%
A novel deep learning architecture using multi-scale temporal cues and spatial-channel refinement improves the robustness of non-contact heart rate measurement from facial videos.
Zhang et al. (2026) studied Heart rate estimation (remote photoplethysmography) (n=85). MDFS and SCOPE modules (proposed rPPG framework) vs. Existing rPPG methods (e.g., RhythmFormer) was evaluated on Mean Absolute Error (MAE) on PURE dataset. The proposed rPPG framework integrating MDFS and SCOPE modules achieved state-of-the-art heart rate estimation on the PURE dataset with a mean absolute error of 0.27 and Pearson correlation of 0.99.