Failed to reproduce results in the paper about abinet HOT 4 CLOSED

Thanks for reporting this issue, can you please provide your training environments and logs, and the accuracy on each dataset?
It is unlikely to have a big gap using this code and I'll try to find out the reason.

from abinet.

@mandal4 Have you resolved this problem, or training details can be provided?
Does the training images you use be cropped directly from original datasets, or after perspective transform?

from abinet.

I found that my training environments are different with official setting

from abinet.

Can you give the differences about the training environments that cause poor performance, which would be helpful.

from abinet.