Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/lisc55/InfoGAN
Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
https://github.com/lisc55/InfoGAN
cnn gan infogan tensorflow tensorlayer
Last synced: about 2 months ago
JSON representation
Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
- Host: GitHub
- URL: https://github.com/lisc55/InfoGAN
- Owner: lisc55
- Created: 2019-10-18T01:45:16.000Z (about 5 years ago)
- Default Branch: master
- Last Pushed: 2023-03-25T01:14:45.000Z (almost 2 years ago)
- Last Synced: 2024-08-02T20:44:31.566Z (5 months ago)
- Topics: cnn, gan, infogan, tensorflow, tensorlayer
- Language: Python
- Homepage:
- Size: 13.1 MB
- Stars: 3
- Watchers: 3
- Forks: 3
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-tensorlayer - InfoGAN
README
# InfoGAN
## InfoGAN Architecture
Tensorlayer implementation of [InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets](https://arxiv.org/abs/1606.03657).
## Results
### MNIST
#### Manipulating the First Continuous Latent Code
Changing 
will rotate the digits:
#### Manipulating the Second Continuous Latent Code
Changing 
will change the width of the digits:
#### Manipulating the Discrete Latent Code (Categorical)
Changing 
will change the type of digits:
#### Random Generation and Loss Plot
G_loss increases steadily after a sufficient number of iterations, showing the discriminator is getting stronger and stronger and indicating the end of training.
### CelebA
#### Manipulating Discrete Latent Code
Azimuth (pose):
Presence or absence of glasses:
Hair color:
Hair quantity:
Lighting:
### Faces
#### Loss Plot
#### Azimuth
#### Random Generation
### Chairs
#### Rotation
## Run
#### MNIST
* Start training using ```python train.py```; this will automatically download the dataset.
* To see the results, execute ```python test.py``` and **input the number of your saved model**.
#### CelebA
+ Set your image folder in `config.py`.
+ Some links for the datasets:
+ [CelebA](https://drive.google.com/drive/folders/0B7EVK8r0v71pWEZsZE9oNnFzTm8)
+ Start training.
```
python train.py
```
#### Faces
* Set your data folder in `config.py`.
* A link for BFM 2009:
* [Basel Face Model](https://faces.dmi.unibas.ch/bfm/main.php?nav=1-0&id=basel_face_model). This should be downloaded before generating data.
* Data is generated using the code in ```data_generator```. Call ```gen_data``` in MATLAB.
* Start training using ```python train.py```.
* To see the results, execute ```python test.py``` and **input the number of your saved model**.
#### Chairs
+ Set your image folder in `data.py`.
+ Some links for the datasets:
+ [Chairs](https://www.di.ens.fr/willow/research/seeing3Dchairs/)
+ Start training.
```
python train.py
```
## References
1. [InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets](https://arxiv.org/abs/1606.03657)
2. [Large-scale CelebFaces Attributes (CelebA) Dataset](http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html)
3. [THE MNIST DATABASE of handwritten digits](http://yann.lecun.com/exdb/mnist/)
4. [Seeing 3D chairs: exemplar part-based 2D-3D alignment using a large dataset of CAD models](https://www.di.ens.fr/willow/research/seeing3Dchairs/)
## Authors
+ [李舒辰 (@lisc55)](https://github.com/lisc55): The experiment on CelebA.
+ [王畅 (@wangchang327)](https://github.com/wangchang327): The experiment on MNIST.
+ [竺沈涵 (@zshCuanNi)](https://github.com/zshCuanNi): The experiment on Faces. Finished by [王畅 (@wangchang327)](https://github.com/wangchang327).
+ [岳鹏云 (@hswd40)](https://github.com/hswd40): The experiment on Chairs.