Thinklab-SJTU/DCL_RetinaNet_Tensorflow

How to calculate Flops and Params?

windyjunfeng opened this issue · 2 comments

Hi, I see you calculate Flops and Params in your paper. I find some codes on the Internet but cant't run those. Could you please share codes about this? Thank you very much!

refer to https://github.com/yangxue0827/RotationDetection/blob/main/tools/train_base.py#L25
But we found that the Flops calculated by this method are not accurate, so we finally used the tool provided by mmdet (by changing the number of output channels of the reg branch of the retinanet method), and updated the relevant content in the final version of the paper.
image

OK, I will try it. Thank you!