Full implementation of YOLOv3 in PyTorch.

[Paper]
[Original Implementation]

• Implement YOLOv3 and darknet53 without original darknet cfg parser.
• It is easy to custom your backbone network. Such as resnet, densenet...
• It is easy to reproduce result of paper using default parameters(~10 epochs).

• pytorch >= 0.4.0
• python >= 3.6.0

git clone https://github.com/BobLiu20/YOLOv3_PyTorch.git
cd YOLOv3_PyTorch
pip3 install -r requirements.txt --user

cd data/
bash get_coco_dataset.sh

1. See weights readme for detail.
2. Download pretrained backbone wegiths from Google Drive or Baidu Drive
3. Move downloaded file darknet53_weights_pytorch.pth to wegihts folder in this project.

1. Review config file training/params.py
2. Replace YOUR_WORKING_DIR to your working directory. Use for save model and tmp file.
3. Adjust your GPU device. see parallels.
4. Adjust other parameters. (keep default to reproduce the results of paper after 10 epochs)

cd training
python training.py params.py

#  please install tensorboard in first
python -m tensorboard.main --logdir=YOUR_WORKING_DIR   

1. See weights readme for detail.
2. Download pretrained yolo3 full wegiths from Google Drive or Baidu Drive
3. Move downloaded file yolov3_weights_pytorch.pth to wegihts folder in this project.

cd evaluate
python eval.py params.py

Please download pretrained weights yolov3_weights_pytorch.pth or use yourself checkpoint.

cd test
python test_images.py params.py

You can got result images in output folder.

Please download pretrained weights yolov3_weights_pytorch.pth or use yourself checkpoint.

cd test
python test_fps.py params.py

• Test in TitanX GPU with different input size and batch size.
• Keep in mind this is a full test in YOLOv3. Not only backbone but also yolo layer and NMS.

@article{yolov3,
	title={YOLOv3: An Incremental Improvement},
	author={Redmon, Joseph and Farhadi, Ali},
	journal = {arXiv},
	year={2018}
}

• darknet
• PyTorch-YOLOv3: Thanks for Evaluate and YOLO loss code