cecret3350/DEA-Net

Issues

• 模型训练问题

  #17 opened a year ago by Baekstar
  9

• 模型验证问题

  #16 opened a year ago by weiMytian
  7

• 模型验证问题

  #29 opened 4 months ago by 17866520451
  9

• 关于重参数化的问题

  #31 opened 3 months ago by ynagyu010603
  0

• 未重参数化的权重

  #30 opened 3 months ago by z-0508
  0

• 单幅雾图去雾问题

  #28 opened 5 months ago by hellohhhhhh
  0

• two device

  #27 opened 5 months ago by thestars-maker
  1

• 评估问题

  #26 opened 6 months ago by wengzp1
  0

• 怎么把自己训练出来的模型用来评估

  #22 opened 10 months ago by ZHB2333
  2

• 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 opened 10 months ago by ouyangxi0925
  1

• CGA fusion

  #25 opened 8 months ago by Fire-friend
  0

• high-level features and low-level features

  #24 opened 8 months ago by Liusir765832
  0

• 关于重参数化reparam.py文件

  #21 opened a year ago by unihuihui
  4

• 您好，请问Fusion模块中，您如何解释(1-W)的作用或目的？

  #20 opened a year ago by planktont
  1

• Backbone和Backbone_train有什么区别，为什么单独分开

  #18 opened a year ago by qqbangbangbang
  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 opened a year ago by xuexisuanfa
  1

• ask for the train code

  #13 opened a year ago by ZeHeru
  1

• About DEConv and train code

  #12 opened a year ago by lyx624
  2

• 作者你好，现在是2023.9.8日

  #11 opened a year ago by zzhlovexuexi
  1

• DEBlock

  #8 opened a year ago by aoe-wyb
  2

• About the training code

  #5 opened a year ago by kkllww
  3

• 模型训练参数

  #19 opened a year ago by ChengchengFU
  7

• 模型验证问题

  #15 opened a year ago by Lxh1992
  3

• DEConv

  #10 opened 2 years ago by Marinerjl
  1

• DEBLOCK

  #9 opened 2 years ago by DFD1
  1

• DEConv

  #7 opened 2 years ago by liux520
  1

• About training settings

  #6 opened 2 years ago by xtxherry
  2

• 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 opened 2 years ago by janice459
  1

• DEBlock

  #3 opened 2 years ago by shangyuu
  1

• A question about DEConv

  #2 opened 2 years ago by hushuaiouc
  1