🎥 Real-time object detection from a Webcam using tiny-YOLO or YOLO with Darkflow (Darknet + tensorflow).
- Python 3+.
Usage of virtualenv is recommended for package library / runtime isolation.
To run the application, please execute the following from the root directory:
-
Setup virtual environment.
-
Install dependencies.
pip3 install -r requirements.lock
- Install Darkflow
source install_darkflow.sh
- Run the application with one of the available models (
tiny-yolo-voc
oryolo
)
python3 -m src [--model MODEL_NAME --device_id DEVICE_ID]
tiny-yolo-voc
is the default model in order to get a better performance w/o GPU. If you want to get better results, then specify theyolo
model.0
is the default device id which is the embedded web cam of your computer. For running the project with another web cam, just specify 1, 2, etc (depends on how many web cams you have).
MIT © YOLO Webcam Object detection