5,322
edits
SinGAN: Learning a Generative Model from a Single Natural Image: Difference between revisions
SinGAN: Learning a Generative Model from a Single Natural Image (view source)
Revision as of 21:05, 13 January 2020
, 13 January 2020→Generator
| Line 30: | Line 30: | ||
The final GAN <math>G_0</math> adds only fine details. | The final GAN <math>G_0</math> adds only fine details. | ||
===Generator=== | ===Generator=== | ||
The use N generators.<br> | The use N generators which they call a hierarchy of patch-GANs.<br> | ||
Each generator consists of 5 convolutional blocks:<br> | Each generator consists of 5 convolutional blocks:<br> | ||
Conv(< | Conv(<math>3 \times 3</math>)-BatchNorm-LeakyReLU.<br> | ||
They use 32 kernels per block at the coarsest scale and increase <math>2 \times</math> every 4 scales. | They use 32 kernels per block at the coarsest scale and increase <math>2 \times</math> every 4 scales. | ||
===Discriminator=== | ===Discriminator=== | ||
The architecture is the same as the generator.<br> | The architecture is the same as the generator.<br> | ||