NAFNet

NAFNet copied to clipboard

arch里 LayerNorm，更接近 InstanceNorm 或 GroupNorm 的变体， mu = x.mean(1, keepdim=True) # shape: (N, 1, H, W) var = (x - mu).pow(2).mean(1, keepdim=True) # shape: (N, 1, H, W) 计算维度：沿 C...

HuiiJi

I would like to ask whether or not it would be possible to retrain the NafNet neural network so that they better fit the specific demands of my work. If...

Leonardo-15

Hi, First of all great project and thank you for your research! I was wondering whether it is possible to upload a NAFNet-REDS-width32 variant model or is there a reason...

kirillmeisser

感谢作者的工作。如果我希望使用我自己的bmp格式的灰度图数据集对网络进行训练，我应该对哪些内容进行修改？

EIEeEss

How much inference time shows?

1

For users who tested this model, how much inference time showed d on your hardware? For me, It showed 4000ms per one 1920*1080 image on RTX 3090 24GB. I wonder...

bongju-kimm

 I ran the script command (test.py) in the path "D:\lunwenfuxian\NAFNet-main", but I got the error shown in the figure.How should I resolve this issue? Thank you for your help...

COS-HSL

  Previously, I changed "create" to "build" in order to prevent this error. Maybe this behavior has something to do with the above error?

COS-HSL

  I ran the script command (test.py) in the path "D:\lunwenfuxian\NAFNet-main", but I got the error shown in the figure.How should I resolve this issue? Do I need to...

COS-HSL

I want to access the model's layers and check the outputs of intermediate layers, but I can't access the layers. Can you help me?

gizemuna

DavidV1600