zhaohengyuan1 / pan Goto Github PK

您好，请问能提供论文图5中各个模型的结果图吗？或者这些应该是在哪里能找到呢？还是说都是需要自己去复现每一篇论文的结果呢？
我想引用对比您们的结果，但不清楚这些经典模型的效果图需要怎么得到？能告诉我下吗？谢谢

Thanks for sharing such a great work, the inference time is also a point to be considered for lightweight. How does the inference time compare to IMDN?

All subprocesses done.
Traceback (most recent call last):
File "extract_subimages.py", line 154, in
main()
File "extract_subimages.py", line 85, in main
data_util._get_paths_from_images(
File "/home/nivetha/Downloads/PAN-master/codes/data/util.py", line 32, in _get_paths_from_images
assert images, '{:s} has no valid image file'.format(path)
AssertionError: /home/Downloads/PAN-master/datasets/DF2K_train/LRx3_sub120 has no valid image file
While running extract_subimages.py code i got error like this,while HR images are created LR images were not created.Flash error like this. How can i solve this.Thanks in advance

你好，在data_scripts/extract_subimages.py中，GT_folder = '/mnt/hyzhao/Documents/datasets/DF2K_train/HR'
LR_folder = '/mnt/hyzhao/Documents/datasets/DF2K_train/LR/X3' 这俩 路径是怎么得来的？
因为 看markdown中 写到关于数据集 只是下载DIV2K和Flickr2K，这两个数据集怎么融合，融合到一起，文件夹结构是什么，可以给解答一下吗

感谢博主这么棒的工作，目前基于DF2K数据集（DIV2K and Flickr2K），我采用./codes/data_scripts/extract_subimages.py file获取x4数据，如下图所示，

基于生成的x4数据，再训练x4的PAN网络，训练29epochs时在Set5上的验证PNSR为30.05，如下图所示，并且训练过程中PNSR增长很慢，请问一下博主当时训练的情况如何呢，是我的配置参数（参数配置如下图2所示）有问题吗？怎么配置才能达到在Set5上PNSR=32.13的指标呢？

感谢您出色的工作和代码。关于模型训练集的选取我有个建议，希望您能在项目的readme中给初学者提示一下。像LatticeNet, RFDN, IMDN等轻量化超分辨模型是仅在DIV2K上面训练的，您提供的代码包含了DF2K。如果初学者直接用DF2K训练网络再与上述模型进行性能对比可能不公平。因此，希望您能在readme中标注一下。再次感谢您的工作。祝好！

I test the pretrained model with your code, but the metric is not match with the paper, the gap is very big(nearly 2dB).
Hoping for your reply.

感谢您在PAN中的工作，我想重新使用您提供的代码训练PAN网络，但是遇到了一些问题，希望您可以解答。
在代码运行到这里的时候就会自动停止，不会继续训练，想问一下这种问题需要怎么解决，期待您的回复！

Thanks for you work. How do you get Params in paper?

Dear zhao：
Hello, I want ask your a question. In your PAN_arch.py ---class PAN in the last two rows, the size of “out” is ([1,3,x, x]) and the size of “ILR” is ([1, 64, x, x]) ,how to add these two tensor. Thanks a lot , look forward your reply