Closed leeqiaogithub closed 4 years ago
我看论文中提到的mask取值是0/1,但是给出的示例中取值是0/255
If I understand correctly, what you want to ask is the value of masks during the training and testing? The mask value is 0/1 (0 for the existing regions and 1 for the corrupted regions)
我看论文中提到的mask取值是0/1,但是给出的示例中取值是0/255