Modifications

Original repo: https://github.com/Tianxiaomo/pytorch-YOLOv4/tree/master
Roboflow version: https://github.com/roboflow/pytorch-YOLOv4
This version is a modified version of the Roboflow version. The modificationa re mainly done to work with pkg version that were installed on Ubuntu 22.0.
Install requirements pip install -r requirements.txt

Notes

Put all your training and validation images in the same directory
Place the annotations files e.g. train.txt and val.txt in the same directory.

Example training command:

python train.py -b 4 -s 1 -l 0.001 -g 0 -pretrained ./checkpoints/Yolov4_epoch15.pth  -classes 2 -dir ./train  -train_label_path ./train/train.txt -val_label_path ./train/val.txt -height 416 -width 416 -epochs 50

Example prediciton command:

python predict.py 80 weights.pth example.jpg classes.txt --resolution 416 --use_cuda 1 --conf_thresh 0.6 --nms_thresh 0.4 --output_img output.jpg

80 is the number of classes. Adjust as needed.

Pytorch-YOLOv4

A minimal PyTorch implementation of YOLOv4.

Paper Yolo v4: https://arxiv.org/abs/2004.10934
Source code:https://github.com/AlexeyAB/darknet
More details: http://pjreddie.com/darknet/yolo/
Inference
Train
- Mocaic

├── README.md
├── dataset.py       dataset
├── demo.py          demo to run pytorch --> tool/darknet2pytorch
├── darknet2onnx.py  tool to convert into onnx --> tool/darknet2pytorch
├── demo_onnx.py     demo to run the converted onnx model
├── models.py        model for pytorch
├── train.py         train models.py
├── cfg.py           cfg.py for train
├── cfg              cfg --> darknet2pytorch
├── data            
├── weight           --> darknet2pytorch
├── tool
│   ├── camera.py           a demo camera
│   ├── coco_annotatin.py       coco dataset generator
│   ├── config.py
│   ├── darknet2pytorch.py
│   ├── region_loss.py
│   ├── utils.py
│   └── yolo_layer.py

0.Weight

0.1 darkent

baidu(https://pan.baidu.com/s/1dAGEW8cm-dqK14TbhhVetA Extraction code:dm5b)
google(https://drive.google.com/open?id=1cewMfusmPjYWbrnuJRuKhPMwRe_b9PaT)

0.2 pytorch

you can use darknet2pytorch to convert it yourself, or download my converted model.

baidu
- yolov4.pth(https://pan.baidu.com/s/1ZroDvoGScDgtE1ja_QqJVw Extraction code:xrq9)
- yolov4.conv.137.pth(https://pan.baidu.com/s/1ovBie4YyVQQoUrC3AY0joA Extraction code:kcel)
google
- yolov4.pth(https://drive.google.com/open?id=1wv_LiFeCRYwtpkqREPeI13-gPELBDwuJ)
- yolov4.conv.137.pth(https://drive.google.com/open?id=1fcbR0bWzYfIEdLJPzOsn4R5mlvR6IQyA)

1.Train

use yolov4 to train your own data

Download weight

Transform data

For coco dataset,you can use tool/coco_annotatin.py.

# train.txt
image_path1 x1,y1,x2,y2,id x1,y1,x2,y2,id x1,y1,x2,y2,id ...
image_path2 x1,y1,x2,y2,id x1,y1,x2,y2,id x1,y1,x2,y2,id ...
...
...

Train

you can set parameters in cfg.py.

 python train.py -g [GPU_ID] -dir [Dataset direction] ...

2.Inference

download model weight https://drive.google.com/open?id=1cewMfusmPjYWbrnuJRuKhPMwRe_b9PaT

python demo.py <cfgFile> <weightFile> <imgFile>

3.Darknet2ONNX

Install onnxruntime
```
pip install onnxruntime
```
Run python script to generate onnx model and run the demo
```
python demo_onnx.py <cfgFile> <weightFile> <imageFile> <batchSize>
```
This script will generate 2 onnx models.
- One is for running the demo (batch_size=1)
- The other one is what you want to generate (batch_size=batchSize)

4.ONNX2Tensorflow

First:Conversion to ONNX

tensorflow >=2.0

1: Thanks:github:https://github.com/onnx/onnx-tensorflow

2: Run git clone https://github.com/onnx/onnx-tensorflow.git && cd onnx-tensorflow Run pip install -e .

Note:Errors will occur when using "pip install onnx-tf", at least for me,it is recommended to use source code installation

Reference:

@article{yolov4,
  title={YOLOv4: YOLOv4: Optimal Speed and Accuracy of Object Detection},
  author={Alexey Bochkovskiy, Chien-Yao Wang, Hong-Yuan Mark Liao},
  journal = {arXiv},
  year={2020}
}

mzahana/yolov4-pytorch