# DCGAN
Using Keras, implemented the face book DCGAN paper.
The pictures below is the result of this DCGAN trained for the simpsons family.
Check out corresponding Kaggle simpson dataset:
[Simpson-Family](https://www.kaggle.com/greg115/image-generator-dcgan-the-simpsons-dataset).
## Architecture
Network architecture by [Radford et al., 2015](https://arxiv.org/abs/1511.0***34).
```
_________________________________________________________________
Model: "Generator"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense_1 (Dense) (None, 16384) 1654784
_________________________________________________________________
re_lu_1 (ReLU) (None, 16384) 0
_________________________________________________________________
reshape_1 (Reshape) (None, 4, 4, 1024) 0
_________________________________________________________________
conv2d_transpose_1 (Conv2DTr (None, 8, 8, 512) 13107200
_________________________________________________________________
batch_normalization_1 (Batch (None, 8, 8, 512) 2048
_________________________________________________________________
re_lu_2 (ReLU) (None, 8, 8, 512) 0
_________________________________________________________________
conv2d_transpose_2 (Conv2DTr (None, 16, 16, 256) 3276800
_________________________________________________________________
batch_normalization_2 (Batch (None, 16, 16, 256) 1024
_________________________________________________________________
re_lu_3 (ReLU) (None, 16, 16, 256) 0
_________________________________________________________________
conv2d_transpose_3 (Conv2DTr (None, 32, 32, 128) 819200
_________________________________________________________________
batch_normalization_3 (Batch (None, 32, 32, 128) 512
_________________________________________________________________
re_lu_4 (ReLU) (None, 32, 32, 128) 0
_________________________________________________________________
conv2d_transpose_4 (Conv2DTr (None, ***, ***, ***) 204800
_________________________________________________________________
batch_normalization_4 (Batch (None, ***, ***, ***) 256
_________________________________________________________________
re_lu_5 (ReLU) (None, ***, ***, ***) 0
_________________________________________________________________
conv2d_transpose_5 (Conv2DTr (None, ***, ***, 32) 51200
_________________________________________________________________
batch_normalization_5 (Batch (None, ***, ***, 32) 128
_________________________________________________________________
re_lu_6 (ReLU) (None, ***, ***, 32) 0
_________________________________________________________________
conv2d_transpose_6 (Conv2DTr (None, ***, ***, 3) 2403
=================================================================
Total params: 19,120,355
Trainable params: 19,118,371
Non-trainable params: 1,***4
_________________________________________________________________
Model: "Discriminator"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_1 (Conv2D) (None, 32, 32, ***) 4800
_________________________________________________________________
leaky_re_lu_1 (LeakyReLU) (None, 32, 32, ***) 0
_________________________________________________________________
conv2d_2 (Conv2D) (None, 16, 16, 128) 204800
_________________________________________________________________
batch_normalization_6 (Batch (None, 16, 16, 128) 512
_________________________________________________________________
leaky_re_lu_2 (LeakyReLU) (None, 16, 16, 128) 0
_________________________________________________________________
conv2d_3 (Conv2D) (None, 8, 8, 256) 819200
_________________________________________________________________
batch_normalization_7 (Batch (None, 8, 8, 256) 1024
_________________________________________________________________
leaky_re_lu_3 (LeakyReLU) (None, 8, 8, 256) 0
_________________________________________________________________
conv2d_4 (Conv2D) (None, 4, 4, 512) 3276800
_________________________________________________________________
batch_normalization_8 (Batch (None, 4, 4, 512) 2048
_________________________________________________________________
leaky_re_lu_4 (LeakyReLU) (None, 4, 4, 512) 0
_________________________________________________________________
flatten_1 (Flatten) (None, 8192) 0
_________________________________________________________________
dense_2 (Dense) (None, 1) 8193
=================================================================
Total params: 4,317,377
Trainable params: 4,315,585
Non-trainable params: 1,792
```