https://github.com/t-almeida/gan-study
Here i present several GAN models in format of notebook implemented with tensorflow using the layers API
https://github.com/t-almeida/gan-study
cgan cnn dcgan gan generative-adversarial-network notebook tensorflow
Last synced: 3 months ago
JSON representation
Here i present several GAN models in format of notebook implemented with tensorflow using the layers API
- Host: GitHub
- URL: https://github.com/t-almeida/gan-study
- Owner: T-Almeida
- Created: 2018-02-02T14:39:08.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2018-08-17T19:04:05.000Z (about 7 years ago)
- Last Synced: 2025-06-22T11:43:00.522Z (4 months ago)
- Topics: cgan, cnn, dcgan, gan, generative-adversarial-network, notebook, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 42.2 MB
- Stars: 5
- Watchers: 2
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# GAN-study
Hi, this repository was created as a student of the master's degree in (computer engineering? i'm not sure that is the correct translation :P), and has the propose to help me understend the topic "generative adversial network". Also the english isn´t my strong point, so i hope i can improve as well.
## GAN models
Here i present several GAN models and basic classifiers (just to compare results) in format of notebook implemented with tensorflow using the layers API
#### Current implementation's
1. GAN (original 2014)
1. Conditional GAN
1. NN for MNIST Classification (simple nn, no dropout or other fancies tecniques, just to have baseline score)
1. Auxiliar classifier GAN
1. CNN for MNIST Classification (just to have baseline score)
1. Deep Convolution GAN
1. Condition Deep Convolution GAN
#### Current/future work
* Condition Deep Convolution GAN
* Semi Supervised GAN (Improved Techniques for Training GANs)
* Information Retrieval GAN
### Material used for current study
* GAN - https://arxiv.org/abs/1406.2661
* Conditional GAN - https://arxiv.org/abs/1411.1784
* Auxiliar classifier GAN - https://arxiv.org/abs/1610.09585
* Deep Convolution GAN - https://arxiv.org/abs/1511.06434
* SSGAN - https://arxiv.org/abs/1606.03498
* Book - Hands-On Machine Learning with Scikit-Learn and TensorFlow (first steps)
* Book code - https://github.com/ageron/handson-ml
* Models code examples - https://github.com/wiseodd/generative-models
### Material for future study
* ICGAN - https://arxiv.org/abs/1611.06355
* EBGAN - https://arxiv.org/abs/1609.03126
* BEGAN - https://blog.heuritech.com/2017/04/11/began-state-of-the-art-generation-of-faces-with-generative-adversarial-networks/
* BEGAN - (paper)
* IRGAN - (paper)
### Requirements
* jupyter notebook
* tensorflow (1.4, 1.5 (soon)) (All the code was run on GPU version (but CPU should work to))
* numpy
* matplotlib