Full-Self Diagnostics (FSD): Physics-Grounded Visual Biomarker Inference from Smartphone Video via Inverse Problems and Operator Learning
Researchers have developed a novel framework called Full-Self Diagnostics (FSD) that can infer physiological biomarkers from short smartphone videos. This system integrates physics-based modeling, information theory, and operator learning to extract data such as spectral, pulse, and micro-expression signals. Empirical validation on over 38,000 videos demonstrated the potential for clinically relevant, non-invasive biomarker inference, with performance improving as more paired biosensor data becomes available. AI
IMPACT This framework could enable widespread, non-invasive health monitoring via consumer devices.