MS-rPPG: Multi-spectral State Space Model for Remote Photoplethysmography in Driver Monitoring Systems

The study presents MS-rPPG, a multi-spectral framework that integrates RGB and near-infrared (NIR) video for remote photoplethysmography (rPPG) aimed at driver health monitoring. It employs a cross-spectral linear modulation (CSLM) strategy for feature integration and introduces MS-Mamba, a state space model that captures long-range temporal dependencies and cross-channel interactions. Evaluated on the MS-Drive and MR-NIRP Car datasets, MS-rPPG demonstrates improved robustness and accuracy in heart rate estimation, which is critical for real-time health monitoring in driving environments.