pix2pix
pix2pix copied to clipboard
phillipi
Image-to-image translation with conditional adversarial nets
My input/output images are grayscale. If I change either input_nc or output_nc to 1, I get an error as the images are loaded as 3 channel images using the "regular"...
Using a very simple model, we can do almost all the tasks of pix2pix  Following is the result using only 179 train samples, no transfer learning, no data augmentation...
I'm attempting to create a model that makes small adjustments to pictures however test.lua produces extremely distorted, cropped, or pixelated images. Is there any way to use a process similar...
I am getting this error when ever I run ./install.sh for torch. I have ubuntu 16.04 with cuda 8.0 and cudnn 5.1 Here is the error that appears on the...
In util.lua, L2 = L2:add(1):mul(50.0) # let L2 ~[0.0, 100.0]. It is reasonable. AB2 = AB2:mul(110.0) # let AB2 ~[-110.0, 110] ? if so, why not [-128,0 127.0]?
``` Resolving people.eecs.berkeley.edu (people.eecs.berkeley.edu)... 128.32.189.73 Connecting to people.eecs.berkeley.edu (people.eecs.berkeley.edu)|128.32.189.73|:443... connected. HTTP request sent, awaiting response... 503 Service Unavailable 2018-02-16 22:01:35 ERROR 503: Service Unavailable. gzip: stdin: unexpected end of file...
Hi, I'm trying to train pix2pix for segmentation. I'm using SUN-RGBD dataset (segmentation in scenes indoor), but my results are very bad. I believe that the problem may be in...
https://github.com/phillipi/pix2pix/blob/master/test.lua#L20 https://github.com/phillipi/pix2pix/blob/master/test.lua#L105-L107 It seems that you test your model only on the default gpu, not consistent with the description of `gpu` option.
If we continue the training, we have to manually do some extra work, it is not quite pleasing. So I create this PR to free this kind of work. Modifications...
Hi, When I start to train, I noticed that apart from the gpu I assigned in train.lua, all other gpus also occupied by 200 to 300 mb memory. Say I...
