DRFNet-Inpainting

code for paper "Degression receptive field network for image inpainting". Simple and efficient implementation.

Prerequisites

• Python 3.7
• NVIDIA GPU + CUDA cuDNN 10.1
• PyTorch 1.8.1

TODO

• Releasing evaluation code.
• Releasing inference codes.
• Releasing pre-trained weights.
• Releasing training codes.

Download Datasets

We use Places2, CelebA-HQ, and Paris Street-View datasets. Liu et al. provides 12k irregular masks as the testing mask.

Run

1. train the model

train.py --dataroot no_use --name Psv_DRFNet --model pix2pixglg --netG1 unet_256 --netD snpatch --gan_modes lsgan --input_nc 4 --no_dropout --direction AtoB --display_id 0

2. test the model

test_and_save_epoch.py --dataroot no_use --name Psv_DRFNet --model pix2pixglg --netG1 unet_256 --gan_mode nogan --input_nc 4 --no_dropout --direction AtoB --gpu_ids 0

pre-trained weights

It's fair to suggest that retrain with your own dataset. You can also use the weight file directly. CelebA-HQ, Paris Street-View, Places2. Please note that for Places2 we randomly selected 2000 images for training and 100 images for testing from each of the first 20 categories.

Citation

@article{MENG2024109397,
  author = {Meng, Jiahao and Liu, Weirong and Shi, Changhong and Li, Zhijun and Liu, Chaorong},
  journal = {Engineering Applications of Artificial Intelligence},
  title = {Degression receptive field network for image inpainting},
  year = {2024},
  volume = {138},
  pages = {109397},
}

Acknowledgments

This code based on LGNet. The evaluation code is borrowed from TFill. Please consider to cite their papers.

@ARTICLE{9730792,
  author={Quan, Weize and Zhang, Ruisong and Zhang, Yong and Li, Zhifeng and Wang, Jue and Yan, Dong-Ming},
  journal={IEEE Transactions on Image Processing}, 
  title={Image Inpainting With Local and Global Refinement}, 
  year={2022},
  volume={31},
  pages={2405-2420}
}
@InProceedings{Zheng_2022_CVPR,
    author    = {Zheng, Chuanxia and Cham, Tat-Jen and Cai, Jianfei and Phung, Dinh},
    title     = {Bridging Global Context Interactions for High-Fidelity Image Completion},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2022},
    pages     = {11512-11522}
}