eriklindernoren / PyTorch-GAN

PyTorch implementations of Generative Adversarial Networks.
MIT License
16.17k stars 4.05k forks source link

modify VGG19 feature map layer in srgan #162

Open Hazarre opened 2 years ago

Hazarre commented 2 years ago

The final srgan in its paper uses the feature map obtained by the 4-th convolution (after activation) before the 5-th maxpooling layer within the VGG19 network.

Also no tanh at the end of conv3 for the generator, just a convolution

baicaiPCX commented 11 months ago

您好,您的邮件小皮已经收到啦!!!