def model_discriminator(input_shape=(1, 28, 28), dropout_rate=0.5): d_input = dim_ordering_input(input_shape, name="input_x") nch = 512 # nch = 128 H = Conv2D( int(nch / 2), (5, 5), strides=(2, 2), padding="same", activation="relu", )(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Conv2D( nch, (5, 5), strides=(2, 2), padding="same", activation="relu", )(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Flatten()(H) H = Dense(int(nch / 2))(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) d_V = Dense(1, activation="sigmoid")(H) return Model(d_input, d_V)
def model_discriminator(input_shape=(1, 92, 92), dropout_rate=0.5): d_input = dim_ordering_input(input_shape, name="input_x") nch = 128 # nch = 128 H = Convolution2D(int(nch / 2), 5, 5, subsample=(2, 2), border_mode='same', activation='relu')(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Convolution2D(nch, 5, 5, subsample=(2, 2), border_mode='same', activation='relu')(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Flatten()(H) H = Dense(int(nch / 2))(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) d_V = Dense(1, activation='sigmoid')(H) return Model(d_input, d_V)
def model_discriminator(input_shape=(1, 28, 28), dropout_rate=0.5): d_input = dim_ordering_input(input_shape, name="input_x") nch = 512 # nch = 128 H = Convolution2D(int(nch / 2), 5, 5, subsample=(2, 2), border_mode='same', activation='relu')(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Convolution2D(nch, 5, 5, subsample=(2, 2), border_mode='same', activation='relu')(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Flatten()(H) H = Dense(int(nch / 2))(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) d_V = Dense(1, activation='sigmoid')(H) return Model(d_input, d_V)
def model_discriminator(input_shape=(1, 28, 28), dropout_rate=0.5): ch_num = 512 d_input = dim_ordering_input(input_shape, name='input_x') H = Conv2D(int(ch_num / 2), (5, 5), strides=(2, 2), padding='same', activation='relu')(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Conv2D(ch_num, (5, 5), strides=(2, 2), padding='same', activation='relu')(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Flatten()(H) H = Dense(int(ch_num / 2))(H) H = LeakyReLU(0.2)(H) d_V = Dense(1, activation='sigmoid')(H) return Model(d_input, d_V)
def model_discriminator(input_shape=(1, 28, 28), dropout_rate=0.5): d_input = dim_ordering_input(input_shape, name="input_x") nch = 512 H = Conv2D(256, 5, 5, subsample=(2, 2), border_mode='same')(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Conv2D(512, 5, 5, subsample=(2, 2), border_mode='same')(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) H = Flatten()(H) H = Dense(256)(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout_rate)(H) d_V = Dense(1, activation='sigmoid')(H) return Model(d_input, d_V)
def discriminator(input_shape=(1, 28, 28), dropout=0.5): d_input = dim_ordering_input(input_shape, name='input_x') c = 512 H = Convolution2D(int(c / 2), kernel_size=5, strides=5, padding='same', activation='relu')(d_input) H = LeakyReLU(0.2)(H) H = Dropout(dropout)(H) H = Convolution2D(c, kernel_size=5, strides=5, padding='same', activation='relu')(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout)(H) H = Flatten()(H) H = Dense(int(c / 2))(H) H = LeakyReLU(0.2)(H) H = Dropout(dropout)(H) d_V = Dense(1, activation='sigmoid')(H) return Model(d_input, d_V)
def model_discriminator(input_shape=(1, 28, 28), dropout_rate=0.5): d_input = dim_ordering_input(input_shape, name="input_x") nch = 512 return Model(d_input, d_V)