神刀安全网

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

DCGAN in Tensorflow

Tensorflow implementation of Deep Convolutional Generative Adversarial Networks which is a stabilize Generative Adversarial Networks. The referenced torch code can be foundhere.

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

To avoid the fast convergence of D (discriminator) network, G (generatior) network is updatesd twice for each D network update which is a different from original paper.

Online Demo

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

link

Prerequisites

Usage

First, download dataset with:

$ mkdir data $ python download.py --datasets celebA 

To train a model with celebA dataset:

$ python main.py --dataset celebA --is_train True --is_crop True 

To test with an existing model:

$ python main.py --dataset celebA --is_crop True 

Or, you can use your own dataset (without central crop) by:

$ mkdir data/DATASET_NAME ... add images to data/DATASET_NAME ... $ python main.py --dataset DATASET_NAME --is_train True $ python main.py --dataset DATASET_NAME 

Results

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

After 6th epoch:

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

After 10th epoch:

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

With asian face dataset (with high noises):

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

More results can be foundhere andhere.

Training details

Details of the loss of Discriminator and Generator (with custom dataset not celebA).

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

Details of the histogram of true and fake result of discriminator (with custom dataset not celebA).

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

Author

Taehoon Kim / @carpedm20

转载本站任何文章请注明:转载至神刀安全网,谢谢神刀安全网 » TensorFlow Implementation of Deep Convolutional Generative Adversarial Networks

分享到:更多 ()

评论 抢沙发

  • 昵称 (必填)
  • 邮箱 (必填)
  • 网址
分享按钮