Background & context
nCyC built a consumer skincare service. The goal: let users measure their own facial-skin condition from a smartphone photo, then receive personalized product recommendations — with no clinical equipment.
Key responsibilities & role
- Skin analysisUsed EfficientNet, VQGAN and OpenCV to detect and quantify facial-skin attributes (pores, wrinkles, oil, moisture) from user photos.
- Data analysisStatistically analyzed self-measured skin data from ~200 users to build a scoring method.
- RecommendationBuilt a score-based recommendation of skincare products from the analysis.
- DeliveryDelivered recommendations through a Google Firebase push-message service.
Tools & knowledge
PythonRSQLEfficientNetVQGANOpenCVFirebaseRecommendation Systems
Results & achievements
→ Enabled a full self-diagnosis + recommendation flow using only a smartphone.
→ Published at ICGHIT 2022: “A Self-Diagnosis System for Measurement of Facial Skin Condition Using Smartphone.”