From e7e8e15d7f89d371f743b71ee14dd7f158814700 Mon Sep 17 00:00:00 2001 From: Jim Martens Date: Tue, 16 Apr 2019 22:08:27 +0200 Subject: [PATCH] Reduce kernel size for encoder/decoder to 5 Signed-off-by: Jim Martens --- src/twomartens/masterthesis/aae/model.py | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/src/twomartens/masterthesis/aae/model.py b/src/twomartens/masterthesis/aae/model.py index a9383bc..1c502d6 100644 --- a/src/twomartens/masterthesis/aae/model.py +++ b/src/twomartens/masterthesis/aae/model.py @@ -46,13 +46,13 @@ class Encoder(keras.Model): def __init__(self, zsize: int) -> None: super().__init__(name='encoder') weight_init = keras.initializers.RandomNormal(mean=0, stddev=0.02) - self.conv1 = keras.layers.Conv2D(filters=zsize * 2, kernel_size=7, strides=2, name='conv1', + self.conv1 = keras.layers.Conv2D(filters=zsize * 2, kernel_size=5, strides=2, name='conv1', padding='same', kernel_initializer=weight_init) self.conv1_a = keras.layers.ReLU() - self.conv2 = keras.layers.Conv2D(filters=zsize * 2, kernel_size=7, strides=2, name='conv2', + self.conv2 = keras.layers.Conv2D(filters=zsize * 2, kernel_size=5, strides=2, name='conv2', padding='same', kernel_initializer=weight_init) self.conv2_a = keras.layers.ReLU() - self.conv3 = keras.layers.Conv2D(filters=zsize, kernel_size=7, strides=2, name='conv3', + self.conv3 = keras.layers.Conv2D(filters=zsize, kernel_size=5, strides=2, name='conv3', padding='same', kernel_initializer=weight_init) self.conv3_a = keras.layers.ReLU() @@ -80,13 +80,13 @@ class Decoder(keras.Model): def __init__(self, channels: int, zsize: int) -> None: super().__init__(name='decoder') weight_init = keras.initializers.RandomNormal(mean=0, stddev=0.02) - self.deconv1 = keras.layers.Conv2DTranspose(filters=zsize * 2, kernel_size=7, strides=2, name='deconv1', + self.deconv1 = keras.layers.Conv2DTranspose(filters=zsize * 2, kernel_size=5, strides=2, name='deconv1', padding='same', kernel_initializer=weight_init) self.deconv1_a = keras.layers.ReLU() - self.deconv2 = keras.layers.Conv2DTranspose(filters=zsize * 2, kernel_size=7, strides=2, name='deconv2', + self.deconv2 = keras.layers.Conv2DTranspose(filters=zsize * 2, kernel_size=5, strides=2, name='deconv2', padding='same', kernel_initializer=weight_init) self.deconv2_a = keras.layers.ReLU() - self.deconv3 = keras.layers.Conv2DTranspose(filters=channels, kernel_size=7, strides=2, name='deconv3', + self.deconv3 = keras.layers.Conv2DTranspose(filters=channels, kernel_size=5, strides=2, name='deconv3', padding='same', kernel_initializer=weight_init) def call(self, inputs: tf.Tensor, **kwargs) -> tf.Tensor: