Minghui Liao

You should also change your network output channel to fit your classes number. the `num_classes` in `modelConfig.py`.

It can do that if you re-train the model.

I can not compile it successfully on my side. Maybe it is sensitive to the environment?

Maybe there are some problems in your "xml" files. Note that the bounding boxes in "xml" files should be the format of "int", not "float".

Most of your labels are difficult?

@jmrichardson Thank you for your instruction!

There is a PR #6 for cudnn6, you can try it.

Ubuntu is suggested. I have no experience with CentOS.

@BIOTONIC The vertexes of the quadrilaterals should be in clockwise.

You just need to load the original image once and rescale it to different scales.