zkl20061823 / jtfn Goto Github PK

In the paper,DRIVE seems to use one-pixel width skeleton for evaluation.Crack200 seems to use four-pixel width (after dilating) for evaluation.How about Crack500?The paper mentioned "In our experiments, the threshold is set to 2".So is it four-pixel width for Crack500 or five-pixel width for Crack500?

thanks for answering!

您好！我按照给出的配置在DRIVE数据集上训练，但是结果并不好，请问是什么原因造成的呢？同时我想问下是如何将骨架拓展到5个像素宽度？可以补充相关代码吗？数据集的训练细节（epoch、lr等）似乎与论文不是完全符合，我应该follow the code吗？期待您的回复，谢谢！

I'm unable to load the pretrained model file that you have provided. Can you please check if the file is correctly uploaded or not?
I'm getting the following error while loading the model:

[enforce fail at inline_container.cc:144] . PytorchStreamReader failed reading zip archive: failed finding central directory

will look forward to your response. Thanks.

作者你好，论文中只给出的分割的评价指标和结果，没有给出边界检测的结果，所以不清楚最后分割出的边界是一个怎样的水平，可以提供一下吗