https://github.com/Hermera/Unsupervised-Attention-guidedImage-to-Image-Translation

A Tensorlayer implementation of the paper "Unsupervised Attention-guided Image-to-Image Translation". (Course Project)
https://github.com/Hermera/Unsupervised-Attention-guidedImage-to-Image-Translation

Last synced: 2 months ago
JSON representation

A Tensorlayer implementation of the paper "Unsupervised Attention-guided Image-to-Image Translation". (Course Project)

• Host: GitHub
• URL: https://github.com/Hermera/Unsupervised-Attention-guidedImage-to-Image-Translation
• Owner: Hermera
• License: mit
• Created: 2019-10-17T06:35:39.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2020-01-10T08:47:59.000Z (about 5 years ago)
• Last Synced: 2024-08-02T20:44:30.833Z (6 months ago)
• Language: Python
• Homepage:
• Size: 11.4 MB
• Stars: 6
• Watchers: 5
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-tensorlayer - Attention CycleGAN - guided Image-to-Image Translation (4. GAN / 1.2 DatasetAPI and TFRecord Examples)

README

        
# Unsupervised Attention-guided Image-to-Image Translation

A Tensorlayer implementation of the paper "[Unsupervised Attention-guided Image-to-Image Translation](https://arxiv.org/pdf/1806.02311.pdf)"[1].

## Network architecture

![network_structure](./imgs/network_structure.png)

## Results on Horse-to-zebra translation

The images within one row are: original input $s$, the attention mask $s_a$, the masked original image $s * s_a$, and the translated image $s'$ respectively. 

- horse2zebra:

![h2z1](./imgs/h2z1.png)

![h2z2](./imgs/h2z2.png)

![h2z3](./imgs/h2z3.png)

- zebra2horse:

    ![z2h1](./imgs/z2h1.png)

    ![z2h2](./imgs/z2h2.png)

    ![z2h3](./imgs/z2h3.png)

## requirements

- Ubuntu 16.04.6 LTS

- Python 3.7

- Tensorflow-gpu 2.0.0

- Tensorlayer 2.1.1

- CUDA 10.0

- anaconda

### Install Tensorflow-gpu 2.0.0 and Tensorlayer with anaconda

```shell

conda create --name project python=3.7

conda install tensorflow-gpu

conda activate project

pip install tensorlayer

```

## usage

### Configure your training and evaluation

Before training and evaluation, you have to configure your dataset and hyperparameters in `config/exp_xx.json`, where `xx` is your experiment name.

You do not need to download the dataset by yourself: the dataset will be automatically downloaded (a zip file) into your folder. See `data_loader.py` for details.

Currently we only support the following datasets:

`horse2zebra, apple2orange, lion2tiger, summer2winter_yosemite `

You can specify your training and testing dataset via editing `"dataset_name"` in `config/exp_xx.json`.

- For example, in training mode, the dataset_name can be `horse2zebra_train`

- In testing mode, the dataset_name can be `horse2zebra_test`

### Training and restoring from the specified checkpoint

To start training, run the `train.sh`:

```shell

bash train.sh

```

You have to modify `train.sh` to specify your configuration file name and where to store all the outputs (checkpoints, tensorboard events, and images).

To restore from the previous checkpoint, run the `restore.sh`

```shell

bash restore.sh

```

Here you have to modify `restore.sh` to specify your previous checkpoint.

### Evaluation

Run `test.sh`  for testing. You can see the results in your log_dir.

```shell

bash test.sh

```

 

## Reference

[1]. [Unsupervised Attention-guided Image-to-Image Translation](https://arxiv.org/pdf/1806.02311.pdf)