this repository only contain the detection module and we don't need the cfg from original YOLOv3, we implement it with pytorch.
This repository is based on the official code of YOLOv3 and pytorch-yolo-v3. There's also an implementation for YOLOv3 already for pytorch, but it uses a config file rather than a normal pytorch approch to defining the network. One of the goals of this repository is to remove the cfg file.
- Python 3.5
- OpenCV
- PyTorch 0.4
- Install PyTorch-0.4.0 by selecting your environment on the website and running the appropriate command.
- Clone this repository
- Compile the nms
- convert yolov3.weights to pytorch
cd YOLOv3_Pytorch
./make.sh
mkdir weights
cd weights
wget https://pjreddie.com/media/files/yolov3.weights
python convert_darknet.py
# we will get the convert_yolov3.pth
python demo.py --images images --save_path ./output --weights ./weights/convert_yolov3.pth