pix2pix
pix2pix copied to clipboard
Image-to-image translation with conditional adversarial nets
Hi, @phillipi I found that the kernel size of generator is 4x4. Why you choose this size? Could you please give me some explaination? Many thanks!
Hi, I'm just getting started with lua so some of my terminology may be wrong here. I'm trying to test the cityscapes_label2image model on the cityscapes dataset. After installing the...
Hi, I want to use pix2pix architecture with the Cifar10 dataset and the problem is the input size. I do not want to resize the images because it is computationally...
Hi, I have some doubts regarding pix2pix. In section 3.3: "At inference time, we run the generator net in exactly the same manner as during the training phase. This differs...
Which images should I use in the evaluation code? I compared the generated images with Ground truth images and generated images with labels but got an error and the result...
Hello! I want to evaluate photo2label on Cityscapes recently(reuse the label-->photo evaluation code) , but when I run the evaluate [command](https://github.com/phillipi/pix2pix#evaluating-labels2photos-on-cityscapes),it'll automatically quit after that: ![image](https://user-images.githubusercontent.com/41133072/53281157-f0ec9200-375e-11e9-8964-6b829a386a95.png) and my result folder...
when training and predicting with your dataset there is not problem and predicts the images correctly but when using a new dataset to predict images it gives a weird result,...
As I can see here in unet like generator network 1st layer and last layer before tanh activation are without BN layer, why? https://github.com/phillipi/pix2pix/blob/master/models.lua#L47 https://github.com/phillipi/pix2pix/blob/master/models.lua#L50 https://github.com/phillipi/pix2pix/blob/master/models.lua#L88 Also by search I...
I am trying to make a new dataset with my own pictures , so I wonder if I need to add bad examples with the label 0 in my dataset.
Hi I have found some difficulties understanding the receptive field in the discriminator. The discriminator takes a 256x256 input and discriminates 30x30 patches with receptive field equal 70x70. I see...