A deep learning system that analyzes tongue images to classify Diabetic vs Non-Diabetic using ResNet50 feature extraction and a custom-built Radial Basis Function Neural Network (RBFNN).
FEATURES:
- ResNet50 feature extraction
- Custom RBF Neural Network implementation
- 5-Fold Cross Validation
- Accuracy: ~95%
- Streamlit Web App
- Confidence scoring
- Tongue image upload interface
TECH STACK: Python, PyTorch, NumPy, Scikit-learn, OpenCV, Streamlit
- Accuracy: 95.50%
- Precision: 91.74%
- Recall (Sensitivity): 100%
- Specificity: 91.00%
- F1 Score: 95.69%
- 5-Fold Cross Validation Accuracy: 95.57% ± 1.10%
1) Clone repository git clone https://github.com/Prantik66/TongueVision-AI.git
2) Move into project folder cd TongueVision-AI
3) Create virtual environment python -m venv venv
4) Activate virtual environment
Windows venv\Scripts\activate
Mac/Linux source venv/bin/activate
5) Install dependencies pip install -r requirements.txt
6) Run Streamlit application streamlit run app/app.py