This project is a Gradio-based web application for detecting Melanoma skin cancer from lesion images using a trained CNN model in TensorFlow. Users can upload an image, and the app classifies it as Benign or Malignant, along with a confidence score. It provides a fast, user-friendly interface for early skin cancer risk assessment.
This is a deep learning-based web application that detects Melanoma (skin cancer) from images of skin lesions using a trained Convolutional Neural Network (CNN) model. The app is built using Gradio and leverages TensorFlow for model inference.
- Upload a skin lesion image
- Predict whether the lesion is Benign or Malignant
- Displays classification result and prediction confidence
- Simple, intuitive web interface using Gradio
- Model file:
melanoma_classification_model.h5 - Input image size: 150x150
- Output: Binary classification (Benign or Malignant)
bash git clone https://github.com/your-username/melanoma-skin-cancer-classifier.git cd melanoma-skin-cancer-classifier
Make sure you are in a virtual environment, then run: bash Copy Edit pip install -r requirements.txt
bash Copy Edit python app.py