SKYN is a cross-platform mobile app for early detection of common skin diseases using deep learning. Built as an undergraduate thesis project, it combines a Flutter frontend, a FastAPI inference backend, and fine-tuned CNN models to give users fast, accessible preliminary skin condition assessments.
Thesis: Dermatological Disease Detection Using a Mobile-Accessible Cloud-Backed Deep Learning System
- Image-based diagnosis with confidence score (upload or capture a photo)
- Detection across five skin conditions — CNNs fine-tuned on 16,549 curated images, 88%+ accuracy
- Diagnosis history tracking
- Role-based access for patients and medical professionals
- Real-time doctor-patient communication
- Firebase-backed auth, storage, and data management
- Frontend: Flutter (Dart), Firebase Auth/Firestore/Storage
- Backend: FastAPI (Python)
- ML: ResNet50 / EfficientNetB0, transfer learning on a curated dermatology dataset
To comply with privacy and security best practices, sensitive Firebase-related information in firebase_config.py, dummy-service-account.json, and firebase.json has been replaced with safe, dummy placeholders.
This project is shared for portfolio and academic purposes only. Do not use, copy, modify, or redistribute this code without my explicit permission.
SKYN is a research and educational tool, not a substitute for professional medical diagnosis. Consult a licensed dermatologist for confirmed diagnosis and treatment.