-
Discriminator
- We set the number of scales (
n_scales_dis
) to be a constant (4) in the paper. This is to verify that a good performance can be achieved for different tasks with a fixed network structure. We found in practice that a reduced number of scales (3) is often discriminative enough and more stabilized, especially for the tasks which require a significant shape deformation. In such case, the 4-th scale, which is the most discriminative one in the default setting, tends to be an overkill. - Reducing the number of base channels (
ch_dis
) is an effective way to accelerate the training process.
- We set the number of scales (
-
Generator
- To improve the inference speed, you may want to reduce the number of base channels (
ch_gen
) or the number of enhanced upsampling layers (n_enhanced_upsample_gen
).
- To improve the inference speed, you may want to reduce the number of base channels (
-
Weak cycle
- The weight for the weak cycle constraint (
cyc_weight
) can be adjusted on a per-task basis. A large value is generally more prohibitive for the shape deformation. On the other hand, it helps to keep the generated image correlated with the source image. - The input of the backward generator is resized by 1/(
resize_factor_gen_bw
). The number of downsampling and upsampling layers of the backward generator is set byn_updownsample_gen_bw
. These two parameters can be adjusted to change the weak cycle to a full resolution forward cycle, or something in between. The effect is somewhat similar to increasing the cycle weight.
- The weight for the weak cycle constraint (
-
Training with a higher resolution
- You may want to adjust several parameters if the input / output resolution is higher than 256x256. Take 512x512 as an example. A good starting point will be setting
img_size
to 512, increasingn_downsample_init_dis
from 2 to 3, and reducingch_dis
,ch_gen
andch_gen_bw
by a factor of two.
- You may want to adjust several parameters if the input / output resolution is higher than 256x256. Take 512x512 as an example. A good starting point will be setting