Graph2Pix: A Graph-Based Image to Image Translation Framework

Installation

Install the dependencies in env.yml

$ conda env create -f env.yml
$ conda activate pix-env

Examples

Train ArtBreeder dataset with Smart Disc Feature enabled and 7 disc images.

$ python train.py --dataroot ./datasets/newbreeder/ --name art_newbreeder_allparents_bs8_disc --no_instance --fineSize 256 --loadSize 256 --label_nc 0 --resize_or_crop resize_and_crop --output_nc 3 --batchSize 8 --mv --smart_disc --num_disc_images 7

Generate images from trained model.

$ python gen_imgs.py --dataroot ./datasets/newbreeder/ --name art_newbreeder_allparents_bs8_disc --no_instance --fineSize 256 --loadSize 256 --label_nc 0 --resize_or_crop resize_and_crop --output_nc 3 --batchSize 8 --mv --smart_disc --num_disc_images 7 --which_epoch 50

FID Calculation

$ python -m pytorch_fid gen_images/art_newbreeder_allparents_bs8_disc/gts gen_images/art_newbreeder_allparents_bs8_disc/generated/

Acknowledgments

The basis of this code is pix2pixHD.

Citation

If you use this code for your research, please cite our paper:

@InProceedings{Gokay_2021_ICCV,
    author    = {Gokay, Dilara and Simsar, Enis and Atici, Efehan and Ahmetoglu, Alper and Yuksel, Atif Emre and Yanardag, Pinar},
    title     = {Graph2Pix: A Graph-Based Image to Image Translation Framework},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
    month     = {October},
    year      = {2021},
    pages     = {2001-2010}
}