darkent常用指令————学习记录

1.直接从头开始训练：darknet.exe detector train .\cfg\projectname.data .\cfg\yolov3_projectname.cfg

训练的过程中backup目录里面每到一定迭代次数，会有模型参数文件保存下来（比如yolov3_projectname_1000.weights）。

2.基于前面训练的模型继续训：darknet.exe detector train .\cfg\projectname.data .\cfg\yolov3_projectname.cfg backup\yolov3_projectname_1000.weights

3.用生成的模型测试图片：darknet.exe detector test .\cfg\projectname.data .\cfg\yolov3_projectname.cfg backup\yolov3_projectname_1000.weights

4.用生成的模型测试视频：darknet.exe detector demo .\cfg\projectname.data .\cfg\yolov3_projectname.cfg backup\yolov3_projectname_1000.weights -i 0 -thresh 0.25 projectname.mp4