cuda >=10.0 python = 3.7

1. extractgallname.py extract the path and name of images from train, gallery and query dataset and saved in train.txt. Commond is python extractgallname.py --root /path/to/dataset/
2. copy train.txt to /darknet/data/
3. under the root of YOLOV4, ./darknet detector test cfg/coco.data cfg/yolov4.cfg yolov4.weights -ext_output -dont_show -out result.json < data/train.txt generates a result.json file in /darknet.
4. under /darknetpath, run json2bike.py to load a json file and generate a bikename.txt in target file. Also can move bike images from root file to target file. Commond is python json2bike.py --data ./result.json --target /target/path/to/save/bike-images/ --root /root/path/