The proposed generator learns both foreground and background attentions. It uses the foreground attention to select from the generated output for the foreground regions, while uses the background attention to maintain the background information from the input image. Please refer to our papers for more details.
AttentionGAN: Unpaired Image-to-Image Translation using Attention-Guided Generative Adversarial Networks.
Hao Tang1, Hong Liu2, Dan Xu3, Philip H.S. Torr3 and Nicu Sebe1.
1University of Trento, Italy, 2Peking University, China, 3University of Oxford, UK.
In TNNLS 2021 & IJCNN 2019 Oral.
The repository offers the official implementation of our paper in PyTorch.
Order: The Learned Attention Masks, The Learned Content Masks, Final Results
Order: The Learned Attention Masks, The Learned Content Masks, Final Results
Order: The Learned Attention Masks, AttentionGAN, StarGAN
Copyright (C) 2019 University of Trento, Italy.
All rights reserved. Licensed under the CC BY-NC-SA 4.0 (Attribution-NonCommercial-ShareAlike 4.0 International)
The code is released for academic research use only. For commercial use, please contact [email protected].
Clone this repo.
git clone https://github.com/Ha0Tang/AttentionGAN
cd AttentionGAN/This code requires PyTorch 0.4.1+ and python 3.6.9+. Please install dependencies by
pip install -r requirements.txt (for pip users)or
./scripts/conda_deps.sh (for Conda users)To reproduce the results reported in the paper, you would need an NVIDIA Tesla V100 with 16G memory.
Download the datasets using the following script. Please cite their paper if you use the data. Try twice if it fails the first time!
sh ./datasets/download_cyclegan_dataset.sh dataset_name
The selfie2anime dataset can be download here.
- Download a dataset using the previous script (e.g., horse2zebra).
- To view training results and loss plots, run
python -m visdom.serverand click the URL http://localhost:8097. - Train a model:
sh ./scripts/train_attentiongan.sh
- To see more intermediate results, check out
./checkpoints/horse2zebra_attentiongan/web/index.html. - How to continue train? Append
--continue_train --epoch_count xxxon the command line. - Test the model:
sh ./scripts/test_attentiongan.sh
- The test results will be saved to a html file here:
./results/horse2zebra_attentiongan/latest_test/index.html.
- You need download a pretrained model (e.g., horse2zebra) with the following script:
sh ./scripts/download_attentiongan_model.sh horse2zebra
- The pretrained model is saved at
./checkpoints/{name}_pretrained/latest_net_G.pth. - Then generate the result using
python test.py --dataroot ./datasets/horse2zebra --name horse2zebra_pretrained --model attention_gan --dataset_mode unaligned --norm instance --phase test --no_dropout --load_size 256 --crop_size 256 --batch_size 1 --gpu_ids 0 --num_test 5000 --epoch latest --saveDisk
The results will be saved at ./results/. Use --results_dir {directory_path_to_save_result} to specify the results directory. Note that if you want to save the intermediate results and have enough disk space, remove --saveDisk on the command line.
- For your own experiments, you might want to specify --netG, --norm, --no_dropout to match the generator architecture of the trained model.
For instance, if you want to run experiments of Selfie to Anime Translation. Usage: replace attention_gan_model.py and networks with the ones in the AttentionGAN-geo folder.
Download data and pretrained model according above instructions.
python test.py --dataroot ./datasets/selfie2anime/ --name selfie2anime_pretrained --model attention_gan --dataset_mode unaligned --norm instance --phase test --no_dropout --load_size 256 --crop_size 256 --batch_size 1 --gpu_ids 0 --num_test 5000 --epoch latest
python train.py --dataroot ./datasets/selfie2anime/ --name selfie2anime_attentiongan --model attention_gan --dataset_mode unaligned --pool_size 50 --no_dropout --norm instance --lambda_A 10 --lambda_B 10 --lambda_identity 0.5 --load_size 286 --crop_size 256 --batch_size 4 --niter 100 --niter_decay 100 --gpu_ids 0 --display_id 0 --display_freq 100 --print_freq 100
python test.py --dataroot ./datasets/selfie2anime/ --name selfie2anime_attentiongan --model attention_gan --dataset_mode unaligned --norm instance --phase test --no_dropout --load_size 256 --crop_size 256 --batch_size 1 --gpu_ids 0 --num_test 5000 --epoch latest
- FID: Official Implementation
- KID or Here: Suggested by UGATIT.
Install Steps:
conda create -n python36 pyhton=3.6 anacondaandpip install --ignore-installed --upgrade tensorflow==1.13.1. If you encounter the issueAttributeError: module 'scipy.misc' has no attribute 'imread', please dopip install scipy==1.1.0.
If you use this code for your research, please cite our papers.
@article{tang2021attentiongan,
title={AttentionGAN: Unpaired Image-to-Image Translation using Attention-Guided Generative Adversarial Networks},
author={Tang, Hao and Liu, Hong and Xu, Dan and Torr, Philip HS and Sebe, Nicu},
journal={IEEE Transactions on Neural Networks and Learning Systems (TNNLS)},
year={2021}
}
@inproceedings{tang2019attention,
title={Attention-Guided Generative Adversarial Networks for Unsupervised Image-to-Image Translation},
author={Tang, Hao and Xu, Dan and Sebe, Nicu and Yan, Yan},
booktitle={International Joint Conference on Neural Networks (IJCNN)},
year={2019}
}
This source code is inspired by CycleGAN, GestureGAN, and SelectionGAN.
If you have any questions/comments/bug reports, feel free to open a github issue or pull a request or e-mail to the author Hao Tang ([email protected]).
I'm always interested in meeting new people and hearing about potential collaborations. If you'd like to work together or get in contact with me, please email [email protected]. Some of our projects are listed here.
Figure out what you like. Try to become the best in the world of it.