densefuse-pytorch
densefuse-pytorch copied to clipboard
Question about mssim
李博你好,
I'm reading your source code and I have two questions:
-
I think there is a little bug in
net.py
, correct me if I am wrong. Innet.py
, I think you want to (but forget to) specify is_last=True for conv5 in decoder. -
What is the benefit of using MS-SSIM instead of original SSIM here? Or MS-SSIM was simply empirically chosen here?
Thanks in advance!
请问代码融合一对图像需要多久
I have the same question as you, I see your implement of the code, This is meaningful for me, have you improve the metrics?