Wildfire Detection Using Deep Learning Models

This repository contains Python implementations for wildfire detection using advanced deep learning models, including Swin V2, DenseNet, GANs, and committee-based approaches. The repository also includes experiments for hyperparameter tuning and feature variance testing.

Table of Contents

• Installation
• Dataset
• Directory Structure
• Experiments
• Usage Instructions
• Contributing
• License

---

Installation

To run the codes, you need to install the required libraries. Use the following commands to install them:

pip install torch torchvision
pip install transformers
pip install matplotlib
pip install scikit-learn
pip install numpy

Dataset

The FLAME Dataset is required to run these experiments. Please download the training and test datasets from the following links:

• Training Data
• Test Data

Code Folder/
├── FLAME_Dataset/
│   ├── Training/
│   └── Test/

Directory Structure

The repository structure should be as follows:

Code Folder/
├── FLAME_Dataset/
│   ├── Training/
│   └── Test/
├── SwinV2-GAN.py
├── SwinPlus_DenseNetEnsemble.py
├── feature_variance.py
├── hyper_parameter.py
├── train_inception.py
├── evaluate_inception.py
├── committee_swin_wildfire_latest-3.py
└── voting_ensemble.py

Experiments

Below are the available experiments in this repository:

1. Swin V2 + GAN

Run the script SwinV2-GAN.py to experiment with the Swin V2 model and GAN-based approach.

2. Swin V2 + DenseNet

Run the script SwinPlus_DenseNetEnsemble.py for the Swin V2 and DenseNet ensemble model.

3. Feature Variance Testing

Run feature_variance.py to analyze feature variance.

4. Hyperparameter Tuning

Run hyper_parameter.py to check the hyperparameter tuning code.

5. Inception V3

• Train the Inception V3 model using train_inception.py.
• Evaluate the trained model using evaluate_inception.py.

6. Committee-Based Approaches

• Sampling-Based Approach: Run committee_swin_wildfire_latest-3.py.
• Model-Based Committee: Run voting_ensemble.py.

Usage Instructions

1. Open your terminal or command line interface.

2. Run the desired experiment using the following command:

   python <experiment_file_name>.py

Example
```bash
python SwinV2-GAN.py

Note:

Depending on your operating system and Python version, you might need to use python3 instead of python.

Contributing

If you’d like to contribute to this project, feel free to fork the repository, make changes, and submit a pull request. Contributions are always welcome!

License

This project is licensed under the MIT License.