jhb86253817/PIPNet

Reproducing NME on test data

Closed this issue · 2 comments

Hi
Thank you for sharing your codes.

I have downloaded one of the dataset (COFW) and tested using available pre-trained models (resnet18). The NME score that I got is 0.08758. If I multiply the score with 100%, it results in 8.758%.

Q: How to produce the same NME score as your paper? In the COFW, it is reported as 3.31%. Thank you.

Hi, what's your PyTorch version? If it's over 1.6, there may be inconsistency for int tensor division, related to line 185 and 190 in functions.py. You can replace "max_ids/tmp_width" by "torch.floor(max_ids/tmp_width)"

My PyTorch version is 1.7. I managed to get the same results by changing line 185 and 190 as you suggested. Thank you!