cecret3350 DEA-Net issues

cecret3350 / DEA-Net

[IEEE TIP 2024] DEA-Net: Single image dehazing based on detail-enhanced convolution and content-guided attention

190 stars 9 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

CGA fusion

#25 Fire-friend opened 1 month ago
0
high-level features and low-level features

#24 Liusir765832 opened 1 month ago
0
训练出现RuntimeError: Tensor for argument #2 'weight' is on CPU, but expected it to be on GPU (while checking arguments for cudnn_batch_norm)的问题

#23 ouyangxi0925 opened 3 months ago
1
怎么把自己训练出来的模型用来评估

#22 ZHB2333 closed 3 months ago
0
关于重参数化reparam.py文件

#21 unihuihui opened 4 months ago
4
您好，请问Fusion模块中，您如何解释(1-W)的作用或目的？

#20 planktont closed 4 months ago
1
模型训练参数

#19 ChengchengFU closed 4 months ago
7
Backbone和Backbone_train有什么区别，为什么单独分开

#18 qqbangbangbang closed 4 months ago
3
模型训练问题

#17 Baekstar opened 5 months ago
6
模型验证问题

#16 weiMytian opened 6 months ago
4
模型验证问题

#15 Lxh1992 closed 5 months ago
3
作者您好，为何用code/train.py 训练后生成的是.pk文件而不是.pth文件，然后使用eval.py运行的时候在加载网络的时候报错

#14 xuexisuanfa closed 4 months ago
1
ask for the train code

#13 ZeHeru closed 4 months ago
1
About DEConv and train code

#12 lyx624 closed 4 months ago
2
作者你好，现在是2023.9.8日

#11 zzhlovexuexi closed 4 months ago
1
DEConv

#10 Marinerjl closed 1 year ago
1
DEBLOCK

#9 DFD1 closed 1 year ago
1
DEBlock

#8 aoe-wyb closed 4 months ago
2
DEConv

#7 liux520 closed 1 year ago
1
About training settings

#6 xtxherry closed 1 year ago
2
About the training code

#5 kkllww closed 4 months ago
3
hello，After reading your code and paper, I found that the DEConv in the code is done by a 3x3 channel invariant convolution, I can't relate to the four difference convolutions and a Vanilla Convolution in the paper, can the author help me to solve this doubt please? Thank you very much.您好！在阅读了您的代码和论文后，发现代码中的DEConv仅由一个3x3通道不变的卷积完成，我无法与论文中四种差分卷积和Vanilla Convolution(香草卷积？普通卷积？PS：原谅我确实不太会翻译)联系起来，请问作者能否帮助我解决这个疑惑吗。非常感谢。真诚.jpg

#4 janice459 closed 1 year ago
1
DEBlock

#3 shangyuu closed 1 year ago
1
A question about DEConv

#2 hushuaiouc closed 1 year ago
1
Fix typo in README.md

#1 eltociear closed 1 year ago
0