FreqPhys: Repurposing Implicit Physiological Frequency Prior for Robust Remote Photoplethysmography
This work addresses the issue of motion artifacts and illumination fluctuations in remote physiological monitoring, which is crucial for applications like healthcare and fitness, but it appears incremental as it builds on existing rPPG methods by incorporating frequency priors.
The paper tackles the problem of robust remote photoplethysmography (rPPG) for contactless physiological monitoring by proposing FreqPhys, a frequency-guided framework that leverages physiological frequency priors to recover signals, achieving significant improvements over state-of-the-art methods on six benchmarks, especially under challenging motion conditions.
Remote photoplethysmography (rPPG) enables contactless physiological monitoring by capturing subtle skin-color variations from facial videos. However, most existing methods predominantly rely on time-domain modeling, making them vulnerable to motion artifacts and illumination fluctuations, where weak physiological clues are easily overwhelmed by noise. To address these challenges, we propose FreqPhys, a frequency-guided rPPG framework that explicitly leverages physiological frequency priors for robust signal recovery. Specifically, FreqPhys first applies a Physiological Bandpass Filtering module to suppress out-of-band interference, and then performs Physiological Spectrum Modulation together with adaptive spectral selection to emphasize pulse-related frequency components while suppress residual in-band noise. A Cross-domain Representation Learning module further fuses these spectral priors with deep time-domain features to capture informative spatial--temporal dependencies. Finally, a frequency-aware conditional diffusion process progressively reconstructs high-fidelity rPPG signals. Extensive experiments on six benchmarks demonstrate that FreqPhys yields significant improvements over state-of-the-art approaches, particularly under challenging motion conditions. It highlights the importance of explicitly modeling physiological frequency priors. The source code will be released.