-
Paper *Deep Plug-And-Play Super-Resolution for Arbitrary Blur Kernels* / DPSR is non-blind model-based SISR method which assumes the blur kernel is known.
https://github.com/cszn/DPSR/issues/2
-
Hi! I liked your approach and wanted to try it for SISR task, however, I am facing issues while loading the weights file when I want to train it on multiple GPUs. I guess the `torch.load()` directly i…
-
-
-
Thank you for doing a great job firstly.
I organized my dataset according to the format of DIV2K, when i trained the model with `python train.py --dataset ./DIV2K_BIN --outdir ./output --profile wds…
-
The extension springs in the MS24586 catalog consistently have an Initial_Tension value that is lower than that required by the default value of SI_Lo_Factor. This presents a challenge for successful…
-
This is my first time learning RefSR. Sun80 or Urban100 sounds like SISR dataset. May I know how to train RefSR code on them? Thanks!
-
很棒的工作!看完模型代码和论文,有几个问题。
1、Relu层的选择,是否有什么经验?比如,大多数SR网络用Relu,DBPN/RBPN用PRelu,ESRGAN和EDVR用LRelu,是基于什么考究?或者说是否以后的实验我什么都别管,直接用LRelu就是最优的?
2、Resblock_noBN中,对Conv1Conv2使用kaiming初始化带来的提升我记得在ESRGAN的补充材料里也出现过,…
-
### Description of defect
The function "GEMALTO_CINTERION_CellularStack::socket_recvfrom_impl" in file "GEMALTO_CINTERION_CellularStack.cpp" implements the lowest layer in a recv call to a TCP so…
-
## Describe the bug
This is not a bug exactly but I need some context regarding how to implement this. I am trying to implement a novel loss function, ref - https://github.com/gfxdisp/mdf. The gist o…