Back to AIBriefs
AnalysisAI ModelsHealth

FCUS-rPPG: Fast-Converging Unsupervised rPPG via Gradient Oscillation Suppression

New unsupervised rPPG method suppresses gradient oscillations for faster convergence without requiring ground-truth physiological annotations. The FCUS-rPPG framework learns BVP representations using consumer-grade cameras.

·
8 days ago
FCUS-rPPG: Fast-Converging Unsupervised rPPG via Gradient Oscillation Suppression — AIBriefs