https://github.com/tomasz-lewicki/fire-detect-nn

Wildfire Detection with Deep Neural Networks
https://github.com/tomasz-lewicki/fire-detect-nn

computer-vision deep-learning fire fire-detection neural-networks

Last synced: about 7 hours ago
JSON representation

Wildfire Detection with Deep Neural Networks

• Host: GitHub
• URL: https://github.com/tomasz-lewicki/fire-detect-nn
• Owner: tomasz-lewicki
• Created: 2019-06-08T17:16:43.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2023-10-03T23:08:00.000Z (over 1 year ago)
• Last Synced: 2024-12-14T08:03:39.591Z (about 1 month ago)
• Topics: computer-vision, deep-learning, fire, fire-detection, neural-networks
• Language: Jupyter Notebook
• Homepage:
• Size: 61.1 MB
• Stars: 37
• Watchers: 3
• Forks: 12
• Open Issues: 5
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Wildfire detection with neural networks

![fire](docs/fire.gif)

## Quickstart

1. Download the repository and pretrained weights

```bash

git clone [email protected]:tomek-l/fire-detect-nn.git

cd fire-detect-nn

wget https://dl.dropbox.com/s/6t17srif65vzqfn/firedetect-densenet121-pretrained.pt --directory-prefix=weights/

```

2. Inference on pretrained weights

```

pip3 install -r requirements.txt

python3 inference.py 

```

3. Training & Testing

- [download the dataset](https://drive.google.com/drive/folders/1j0Smp1ALUdZiAt4a3qEFH_85oMc17vsV?usp=sharing).

- unzip the folder to `data/fire-detect-nn-public-combo`

```bash

python3 train.py # training 

python3 train-with-gradcam.py # for the model with gradcam (for the heatmap output)

python3 test.py # evaluation

```

For the GradCAM heatmap generation use jupyter notebook and navigate to `inference-video.ipynb`