diff options
| -rwxr-xr-x | cdcgan.py | 66 |
1 files changed, 34 insertions, 32 deletions
@@ -1,8 +1,5 @@ from __future__ import print_function, division -import tensorflow as keras - -import tensorflow as tf -import tensorflow.keras as keras +import keras from keras.datasets import mnist from keras.layers import Input, Dense, Reshape, Flatten, Dropout, multiply from keras.layers import BatchNormalization, Embedding, Activation, ZeroPadding2D @@ -39,11 +36,12 @@ class CDCGAN(): optimizer=optimizer, metrics=['accuracy']) - # Build the generator - self.generator = self.build_generator() - noise = Input(shape=(self.latent_dim,)) label = Input(shape=(1,)) + + # Build the generator + self.generator = self.build_generator(noise, label) + img = self.generator([noise, label]) # For the combined model we will only train the generator @@ -59,38 +57,44 @@ class CDCGAN(): self.combined.compile(loss=['binary_crossentropy'], optimizer=optimizer) - def build_generator(self): + def build_generator(self, noise_in, label_in): + noise = Dense(7 * 7 * 256)(noise_in) + noise = Reshape(target_shape=(7, 7, 256))(noise) + noise = Conv2DTranspose(256, kernel_size=3, padding="same")(noise) + noise = BatchNormalization()(noise) + noise = Activation("relu")(noise) - model = Sequential() + label = Flatten()(Embedding(self.num_classes, self.latent_dim)(label_in)) + label = Dense(7 * 7 * 256)(label) + label = Reshape(target_shape=(7, 7, 256))(label) + label = Conv2DTranspose(256, kernel_size=3, padding="same")(label) + label = BatchNormalization()(label) + label = Activation("relu")(label) - model.add(Dense(128 * 7 * 7, activation="relu", input_dim=self.latent_dim)) - model.add(Reshape((7, 7, 128))) + # Combine the two - model.add(Conv2DTranspose(256, kernel_size=3, padding="same", strides=(2,2))) - model.add(BatchNormalization()) - model.add(Activation("relu")) + x = keras.layers.Concatenate()([noise, label]) - model.add(Conv2DTranspose(128, kernel_size=3, padding="same", strides=(2,2))) - model.add(BatchNormalization()) - model.add(Activation("relu")) + x = Conv2DTranspose(256, kernel_size=3, padding="same")(x) + x = BatchNormalization()(x) + x = Activation("relu")(x) - model.add(Conv2DTranspose(64, kernel_size=3, padding="same")) - model.add(BatchNormalization()) - model.add(Activation("relu")) + x = Conv2DTranspose(128, kernel_size=3, padding="same", strides=(2,2))(x) + x = BatchNormalization()(x) + x = Activation("relu")(x) + + x = Conv2DTranspose(64, kernel_size=3, padding="same", strides=(2,2))(x) + x = BatchNormalization()(x) + x = Activation("relu")(x) - model.add(Conv2DTranspose(1, kernel_size=3, padding="same")) - model.add(Activation("tanh")) + x = (Conv2DTranspose(1, kernel_size=3, padding="same"))(x) + x = Activation("tanh")(x) + model = Model([noise_in, label_in], outputs=x) - noise = Input(shape=(self.latent_dim,)) - label = Input(shape=(1,), dtype='int32') - label_embedding = Flatten()(Embedding(self.num_classes, self.latent_dim)(label)) - model_input = multiply([noise, label_embedding]) - img = model(model_input) - - #model.summary() + model.summary() - return Model([noise, label], img) + return model def build_discriminator(self): @@ -242,8 +246,6 @@ class CDCGAN(): return train_data, test_data, val_data, labels_train, labels_test, labels_val -''' if __name__ == '__main__': cdcgan = CDCGAN() cdcgan.train(epochs=4000, batch_size=32) -''' |