Non-stationary texture synthesis using adversarial expansions

This is my newer version code adapted from the official one Non-stationary texture synthesis using adversarial expansions.

What's New?

• The code can be run on Windows 10, Python 3.7.9, Pytorch 1.6.0, CUDA 10.1.
• Fixed HTML stuff(not used)

Run

Download VGG

Download VGG model from here and put it inside models folder.

Train

python train.py --dataroot ./datasets/half/5 --name 5_half_style_14x14 --use_style --no_flip --no_lsgan --padding_type replicate --model half_style --which_model_netG resnet_2x_6blocks --which_model_netD n_layers --n_layers_D 4 --which_direction AtoB --lambda_A 100 --dataset_mode half_crop --norm batch --pool_size 0 --resize_or_crop no --niter_decay 50000 --niter 50000 --save_epoch_freq 2000

My Results

Input	Synthesized

Acknowledgements

Official Code