A TensorFlow implementation of YOLOv3.
Reference: YOLOv3: An Incremental Improvement
This was developed with the following environment:
- Ubuntu 16.04
- CUDA Toolkit 9.0
- cuDNN SDK v7.0
- NVIDIA GPU Driver 390.59
- Python 3.5
- All Python packages listed in
requirements.txt
The following assumes that you have already set up a virtual environment for Python 3.5. Also, the setup.sh
script has only been tested on Ubuntu 16.04, and not other operating systems.
Follow the instructions for either the CPU or GPU setup, depending on what you want to use.
Run ./setup.sh cpu
.
If you want to use your GPU, follow this link to install CUDA and cuDNN for TensorFlow on Linux.
Then, run ./setup.sh
.
To run YOLOv3, type the following in the terminal:
python main.py <path_to_image_or_video>
To run the test cases, type the following in the terminal:
python run_tests.py