MirrorGAN-master
所属分类:人工智能/神经网络/深度学习
开发工具:Python
文件大小:188KB
下载次数:1
上传日期:2021-02-02 12:54:51
上 传 者:
imagepingjia
说明: Pytorch implementation for Paper MirrorGAN: Learning Text-to-image Generation by Redescription by Tingting Qiao, Jing Zhang, Duanqing Xu, Dacheng Tao. (The work was performed when Tingting Qiao was a visiting student at UBTECH Sydney AI Centre in the School of Computer Science, FEIT, the University of Sydney).
文件列表:
.idea (0, 2019-07-22)
.idea\MirrorGAN.iml (398, 2019-07-22)
.idea\encodings.xml (135, 2019-07-22)
.idea\misc.xml (288, 2019-07-22)
.idea\modules.xml (270, 2019-07-22)
.idea\vcs.xml (180, 2019-07-22)
GLAttention.py (4593, 2019-07-22)
cfg (0, 2019-07-22)
cfg\__init__.py (0, 2019-07-22)
cfg\config.py (2820, 2019-07-22)
cfg\eval_bird.yml (476, 2019-07-22)
cfg\train_bird.yml (866, 2019-07-22)
datasets.py (11129, 2019-07-22)
do_test.sh (48, 2019-07-22)
do_train.sh (49, 2019-07-22)
images (0, 2019-07-22)
images\framework.jpg (273002, 2019-07-22)
main.py (5098, 2019-07-22)
miscc (0, 2019-07-22)
miscc\__init__.py (70, 2019-07-22)
miscc\losses.py (7793, 2019-07-22)
miscc\utils.py (10954, 2019-07-22)
model.py (28148, 2019-07-22)
pretrain_DAMSM.py (10745, 2019-07-22)
test.py (422, 2019-07-22)
trainer.py (21646, 2019-07-22)
# MirrorGAN
Pytorch implementation for Paper [MirrorGAN: Learning Text-to-image Generation by Redescription](https://arxiv.org/abs/1903.05854) by Tingting Qiao, Jing Zhang, Duanqing Xu, Dacheng Tao. (The work was performed when Tingting Qiao was a visiting student at UBTECH Sydney AI Centre in the School of Computer Science, FEIT, the University of Sydney).
![image](images/framework.jpg)
## Getting Started
### Installation
- Install PyTorch and dependencies from http://pytorch.org
- Install Torch vision from the source.
- Clone this repo:
```bash
git clone https://github.com/qiaott/MirrorGAN.git
cd MirrorGAN
```
- Download our preprocessed data from [here](https://drive.google.com/file/d/1CuW5ognTSkNbyx9TWoUFrgwqxZNk1cl0/view?usp=sharing).
- The STEM was pretrained using the code provided [here](https://github.com/taoxugit/AttnGAN)
- The STREAM was pretrained using the code provided [here](https://github.com/yunjey/pytorch-tutorial/tree/master/tutorials/03-advanced/image_captioning).
### Train/Test
After obtaining the pretrained STEM and STREAM modules, we can train the text2image model.
- Train a model:
```bash
./do_train.sh
```
- Test a model:
```bash
./do_test.sh
```
## Citation
If you use this code for your research, please cite our paper.
```bash
@article{qiao2019mirrorgan,
title={MirrorGAN: Learning Text-to-image Generation by Redescription},
author={Qiao, Tingting and Zhang, Jing and Xu, Duanqing and Tao, Dacheng},
journal={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
year={2019}
}
```
近期下载者:
相关文件:
收藏者: