Tamar Rott Shaham
Tamar Rott Shaham
for SR, SinGAN is trained with an upsampling factor of 2^(1/3) between levels, so to get an SR factor of x4 for example, will need to use the upsampling process...
We found that with initial padding we get higher variability on the edges of the generated images, than the case of using zero padding at each conv layer. See the...
Your modification sounds correct. Please also check if you need to change anything within the imresize function.
Seems like you are trying to take the mean of a nan value. Make sure you input the correct image path.
If you wish to, you can play with all the hyper parameters (e.g. number of channels, number of layers, etc.) and monitor the learning curves.
Thank you Clef! Would be happy to hear if all tests worked fine. Should I push this to the main brunch?
This code seems to do a single D step per G step (in groups of three steps), and is therefore not identical to the original implementation (were 3 D steps...
you can create the naively pasted version and the mask manually, using photoshop/paint.net or similar programs
You should use higher value for harmonization_start_scale. You can play with this parameter until you'll get the desired result. On Fri, Nov 22, 2019, 04:26 Johnson yue wrote: > Hi,...
This is the result I get with start_scale=6:  Did you change anything within the function?