Yuhui Ma

那应该如何修改代码呢

好的，我看看。

Sorry, I am not the author. I don't even know how to run it on my dataset. The author could be busy and may not response to you.

I found you have commented in every issue. Why?

@BeyondHeaven Yes, but I haven't run the code. Please give me your e-mail, and then I will share the code with you.

The code is only able to split two categories - foreground and background now. I apply the project on OCT images to complete the task of choroidal segmentation. So that...

Hello Tahlil Ahmed Chowdhury, It's not clear what's wrong with you. Maybe you can save the segmentation maps during training and observe them combined with loss and iou. You can...

> The iou actually going to the opposite direction. Its decreasing. First couple of iteration we can see that it decreased to nearly 0. >  >...

You should compute softmax of these values. So you can get pixel values from 0 to 1.

It's incredible that no foreground is segmented while loss is decreased. Let me see the pw_softmaxwithloss_2d and focal_loss. They shouldn't have any problems. It's so confusing. Maybe you could observe...