Chong Mou

你好，那个train_block是为了消除压缩感知过程中的块效应问题

They are not two steps but two versions in our paper.

Thank you for your constructive suggestion! This modification is correct and effective!

Yes, DGUNet+ takes more training time.

About 6 days with 4 V100 GPUs.

Hi, Phase 1 trains the network with MSE loss, and Phase 2 is trained with the GAN loss with the initial parameters from Phase 1. About 7 days to complete...

Sorry for the misunderstanding. We did some debugging before releasing the code. We found that it is unnecessary to train the net stage for too long, so we shortened the...

Hi, thank you for your attention. You need to install basicsr first and then utilize this function.

Thanks for your reply! Does it mean the l_forw_ce = self.train_opt['lambda_ce_forw'] * torch.sum(z**2) / z.shape[0] is only an assistant component rather than a strict constraint of gaussian distributions?

Thanks for your reply! Does it mean the l_forw_ce = self.train_opt['lambda_ce_forw'] * torch.sum(z**2) / z.shape[0] is only an assistant component rather than a strict constraint of gaussian distributions?