ShareHealth

MAJOR UPDATE: Face to Health   - Added first-launch medical disclaimer before health features are used. - Added in-app medical sources/citations links for health insights. - Improved review compliance and reliability • NEW: Face to Health — predict health metrics from facial analysis using on-device AI • NEW: MediaPipe integration for precise 478-point facial landmark detection • NEW: Personal ML model training with leave-one-out cross-validation • NEW: Face History view to track your captures over time • NEW: Import face captures from folders with automatic metrics calculation • NEW: Sleep Score calculation from Apple Health data   • NEW: Random Forest models — capture non-linear face-to-health patterns that linear regression might miss • NEW: Model type selector — switch between Linear Regression and Random Forest with one tap; your preference saves automatically • NEW: Enhanced feature importance — see which of 24 facial features matter most for each health prediction • IMPROVED: Model snapshots now preserve your chosen algorithm • IMPROVED: Cross-validation analysis with clearer R correlation, MAE, and RMSE metrics • IMPROVED: Health data export with richer formatting • IMPROVED: Streamlined onboarding experience • FIXED: Various stability and performance improvements Why two algorithms? Linear Regression excels at clear, interpretable correlations. Random Forest (50 decision trees) can detect threshold effects and complex interactions between facial features. Try both to see which works better for your unique physiology. All processing remains 100% on-device.