csZcWu/GKMNet

Issues

• if you have the plan to open the train code

  #1 opened a year ago by TimZhang001
  0

• 开源复现与论文结果相比较，差距很大。测试选定的哪个epoch？能否公布模型详细参数。

  #4 opened 2 years ago by yangt1013
  10

• The training code produces much worse performance than the paper

  #3 opened 3 years ago by SkyeLu
  2

• 其实很多应用是只需要增大景深，而不是全深度清晰。所以是否可以控制去模糊卷积核的参数来进行限制呢？

  #2 opened 3 years ago by TimZhang001
  1