Longguang Wang

Hi @zhangqizky , many thanks for your interest in our work. We plan to release the code before the conference.

Hi @YuanfeiHuang, thanks for you interest in our work. We have uploaded pretrained models for x3 and x4 SR.

@cheun726，可以的，欢迎尝试😀

几个sh脚本都可以在windows下运行，请参考readme配置好环境、准备好数据集并修改sh文件中的路径即可。

Hi @Lewis0427, please leave your email address.

@jzw0707 你好，能提供一下这个测试结果的参数设置么？

我在我这边IDE里直接run的结果是没问题。看你的结果值应该是pre-trained model没有加载，请问有对代码其他地方进行修改么？ 

Hi @Aacrobat, 感谢对我们工作的关注。无监督是指在学习degradation representation的时候，不需要groundtruth degradation提供监督。比如在一些真实采集的HR-LR数据上（比如RealSR数据集），groundtruth degradation是很难获得的，导致已有的一些盲超分方法无法进行训练，而我们的DASR可以较好地解决这个问题。

Hi @aupendu , many thanks for your interest in our work. We plan to release the code before the conference.

Hi @choieq, thanks for your interest in our work. In our implementation, we generate LR images online in the dataloader using random degradations (see [here](https://github.com/LongguangWang/DASR/blob/main/trainer.py#L155)). So we crop HR image...