pytorch-SRResNet
pytorch-SRResNet copied to clipboard
twtygqyy
pytorch implementation for Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network arXiv:1609.04802
Hello, I want to fine-tune/train the pre-trained model for my dataset.... So what learning rate should I use ??
Have anyone tried to test this work on Google Colab? Failed to set up matlab.engine.
why dose the model generate output image with artifacts when combined with the GAN loss ?
Regarding your SRResNet implementation (the generator), I was wondering why you used a LeakyReLU instead of a PReLU with a trainable "a". Reading the paper and seing your implementation raised...
Hi , I'm training srresnet on DIV2K from scratch(no pretrained) at lr: e-4 , and it is decayed by 0.1 for step=200, but I'm having MSE loss(at 402 epoch) as:-...
Shouldn't the repo be called pytorch-SRGAN instead of pytorch-SRResNet? Anyway, the same group mention using [orthogonal weight initialisation](http://pytorch.org/docs/master/nn.html#torch.nn.init.orthogonal) in the [ESPCN](https://arxiv.org/abs/1611.05250) paper released roughly at the same time period -...
Hi, @twtygqyy Thank you very much for the work. I read your code, but I feel confused about your annotation in [srresnet.py](https://github.com/twtygqyy/pytorch-SRResNet/blob/master/srresnet.py). ``` class _NetD(nn.Module): def __init__(self): super(_NetD, self).__init__() self.features...
I want to calculate PSNR on Y channel. I have got `im_gt_y` and `im_b_y` through the mat format files. ``` im_gt_y = sio.loadmat("output/" + opt.image + ".mat")['im_gt'] im_b_y = sio.loadmat("output/"...
