hi,thanks for your work, but mAp is too bad, how can i improve the result?

eric612 / Caffe-YOLOv3-Windows

A windows caffe implementation of YOLO detection network

Other

212 stars 128 forks source link

hi,thanks for your work, but mAp is too bad, how can i improve the result? #1

Closed aa12356jm closed 6 years ago

eric612 commented 6 years ago

I got same problem with issue The region loss output are same as caffe , so I guess problem is data augmentation, where the original project has jitter,random size,image distortion.