DermaDetectAI

DermaDetectAI is a Flask-based application developed to detect various skin diseases using deep learning models. This project was created as part of a college initiative by Ranit Kumar Manik, Mohammad, Sayak Bal, and Partha Sarathi Manna. It features three distinct models, each trained on different datasets using PyTorch to identify 5, 10, and 23 skin diseases, respectively.

Table of Contents

• Pre-trained Models
• Setup Instructions
  • Install Dependencies
  • Running the Application
• Using the Pre-trained Models
• Training the Models
• Project Structure
• Contributing
• License

Pre-trained Models

The repository includes pre-trained models for skin disease detection:

1. Model 1: Detects 5 diseases. Trained on a ~69MB dataset with 98% validation accuracy.
2. Model 2: Detects 10 diseases. Trained on a ~2GB dataset with 85% validation accuracy.
3. Model 3: Detects 23 diseases. Trained on a ~6GB dataset with 45% validation accuracy.

Setup Instructions

Install Dependencies

Each model has its own requirements.txt file. To install the dependencies for a specific model, navigate to the respective model directory and run:

pip install -r requirements.txt

Running the Application

To start the Flask application for a specific model, navigate to its directory and execute:

python app.py

The Flask server will start, and you can access the application at http://127.0.0.1:5000. Use the web interface to upload an image and receive disease predictions.

Using the Pre-trained Models

The pre-trained models are included in the repository, allowing you to use them directly without additional training.

Training the Models

To train the models from scratch, navigate to the src directory of the respective model and run main.py. Ensure that you have the dataset in the appropriate directory and adjust the num_classes parameter according to your dataset's number of classes.

python src/main.py

Note

This project is configured to utilize NVIDIA GPUs for faster training and inference. Make sure you have the necessary NVIDIA drivers, CUDA toolkit, and the GPU version of PyTorch installed.
For GPU setup instructions, refer to the CUDA Installation Guide. For PyTorch installation guidance, visit the PyTorch Installation Page.

Project Structure

Here’s an overview of the project structure:

DermaDetectAI/
├── LICENSE
├── README.md
├── model-X/
│   ├── app.py
│   ├── models/
│   │   └── skin_disease_model.pth
│   ├── requirements.txt
│   ├── src/
│   │   └── main.py
│   ├── templates/
│   │   ├── result.html
│   │   └── upload.html
│   └── uploads/
│       └── [user_uploaded_files]
└── [other_files_and_directories]

For more details, refer to the Project Structure Documentation.

Contributing

We welcome contributions to this project! To contribute, please follow these steps:

1. Fork the repository: Click the "Fork" button at the top right of this page to create a copy of the repository under your GitHub account.
2. Clone the repository: Clone your forked repository to your local machine:

   git clone https://github.com/RanitManik/DermaDetectAI.git

3. Create a new branch: Create a new branch for your feature or bug fix:

   git checkout -b feature-or-bugfix-name

4. Make your changes: Implement your changes to the codebase.
5. Commit your changes: Commit your changes with a descriptive message:

   git commit -m "Description of your changes"

6. Push to your branch: Push your changes to your forked repository:

   git push origin feature-or-bugfix-name

7. Create a Pull Request: Open a pull request from your forked repository’s branch to the main branch of the original repository.

For detailed contribution guidelines, please refer to the CONTRIBUTING.md file.

License

This project is licensed under the MIT License. See the LICENSE file for details.