YOLOV2 Car Detection Project

This project applies YOLOV2 for car detection using the Keras version from allanzelener/YAD2K. The goal is to use YOLO in the field of an autonomous driving applications. The car detection is performed on a dataset provided by drive.ai.

Requirements

• numpy
• jupyterlab
• matplotlib
• imageio
• imageio[ffmpeg]
• scipy
• Pillow
• ipython
• nodejs-bin
• jupyter_contrib_nbextensions
• tensorflow
• pandas The requirements are also listed in requirements.txt.

Installation

1. Install the required packages:
   These packages can be installed by running the following command:

# Load pipenv env
pipenv shell
# Install packages
pipenv install --requirements "requirements.txt"
# Install ipython in jupyter lab
jupyter labextension install @jupyter-widgets/jupyterlab-manager
jupyter labextension install jupyter-matplotlib
# Fixe jupyter_nbextensions_configurator error
jupyter nbextensions_configurator enable --user

2. Load weights:
   download yolo.weights and yolov2.cfg and put them in model_data, and then, transform them to yolo.h5

wget http://pjreddie.com/media/files/yolo.weights
wget https://raw.githubusercontent.com/pjreddie/darknet/master/cfg/yolov2.cfg
python yad2k.py model_data\yolov2.cfg model_data/yolo.weights model_data/yolo.h5

3. Launch JupyterLab:

jupyter-lab

4. Open the car_detection.ipynb notebook and run the cells to perform car detection on the provided dataset.

Project Structure

The project has the following tree structure:

.
├── font
│   └── FiraMono-Medium.otf
├── images
│   ├── 001.jpg
│   ├── ...
│   └── 0120.jpg
├── model_data
│   ├── _anchors.txt
│   ├── coco_classes.txt
│   ├── pascal_classes.txt
│   ├── saved_model.pb
│   └── yolo_anchors.txt
├── yad2k
│   ├── ...
│   └── ...
├── yad2k.py
├── Pipfile
├── Pipfile.lock
├── car_detection.ipynb
├── README.md
└── LICENSE.md

License

This project is licensed under the MIT License - see the LICENSE.md file for details.