Build Functional Model, remove subclassed model

Functional models are way easier to work with,
and I don't need any advanced features that would
require model subclassing

fast_depth_functional.py

@@ -5,59 +5,53 @@ Functional version of fastdepth model. Note that this doesn't work at the moment
 '''
 
 
-def _depthwise_conv_block(inputs,
+def FDDepthwiseBlock(inputs,
-                          pointwise_conv_filters,
+                     out_channels,
-                          depth_multiplier=1,
+                     block_id=1):
-                          strides=(1, 1),
+    x = keras.layers.DepthwiseConv2D(5, padding='same')(inputs)
-                          block_id=1):
+    x = keras.layers.BatchNormalization()(x)
-    channel_axis = 1 if keras.backend.image_data_format() == 'channels_first' else -1
+    x = keras.layers.ReLU(6.)(x)
-    pointwise_conv_filters = int(pointwise_conv_filters)
+    x = keras.layers.Conv2D(out_channels, 1, padding='same')(x)
-
+    x = keras.layers.BatchNormalization()(x)
-    if strides == (1, 1):
-        x = inputs
-    else:
-        x = keras.layers.ZeroPadding2D(((0, 1), (0, 1)), name='conv_pad_%d' % block_id)(
-            inputs)
-    x = keras.layers.DepthwiseConv2D((3, 3),
-                                     padding='same' if strides == (1, 1) else 'valid',
-                                     depth_multiplier=depth_multiplier,
-                                     strides=strides,
-                                     use_bias=False,
-                                     name='conv_dw_%d' % block_id)(
-        x)
-    x = keras.layers.BatchNormalization(
-        axis=channel_axis, name='conv_dw_%d_bn' % block_id)(
-        x)
-    x = keras.layers.ReLU(6., name='conv_dw_%d_relu' % block_id)(x)
-
-    x = keras.layers.Conv2D(
-        pointwise_conv_filters, (1, 1),
-        padding='same',
-        use_bias=False,
-        strides=(1, 1),
-        name='conv_pw_%d' % block_id)(
-        x)
-    x = keras.layers.BatchNormalization(
-        axis=channel_axis, name='conv_pw_%d_bn' % block_id)(
-        x)
     return keras.layers.ReLU(6., name='conv_pw_%d_relu' % block_id)(x)
 
 
 def make_fastdepth_functional():
     # This doesn't work, at least right now...
-    mobilenet = keras.applications.MobileNet(include_top=False)
     input = keras.layers.Input(shape=(224, 224, 3))
-    x = mobilenet(input)
+    x = input
-    x = _depthwise_conv_block(x, 512, block_id=14)
+    mobilenet = keras.applications.MobileNet(include_top=False, weights=None)
-    x = _depthwise_conv_block(x, 256, block_id=15)
+    for layer in mobilenet.layers:
-    x = keras.layers.Add()([x, mobilenet.get_layer('conv_pw_5_relu').output])
+        x = layer(x)
-    x = _depthwise_conv_block(x, 128, block_id=16)
+        if layer.name == 'conv_pw_5_relu':
-    x = keras.layers.Add()([x, mobilenet.get_layer('conv_pw_3_relu').output])
+            conv5 = x
-    x = _depthwise_conv_block(x, 64, block_id=17)
+        elif layer.name == 'conv_pw_3_relu':
-    x = keras.layers.Add()([x, mobilenet.get_layer('conv_pw_1_relu').output])
+            conv3 = x
-    x = _depthwise_conv_block(x, 32, block_id=18)
+        elif layer.name == 'conv_pw_1_relu':
+            conv1 = x
 
-    x = keras.layers.Conv2D(1, 1)(x)
+    x = FDDepthwiseBlock(x, 512, block_id=14)
+    # TODO: Bilinear interpolation
+    #     x = keras.layers.experimental.preprocessing.Resizing(14, 14)
+    # Nearest neighbour interpolation, used by fast depth paper
+    x = keras.layers.experimental.preprocessing.Resizing(14, 14, interpolation='nearest')(x)
+    x = FDDepthwiseBlock(x, 256, block_id=15)
+    x = keras.layers.experimental.preprocessing.Resizing(28, 28, interpolation='nearest')(x)
+    x = keras.layers.Add()([x, conv5])
+    x = FDDepthwiseBlock(x, 128, block_id=16)
+    x = keras.layers.experimental.preprocessing.Resizing(56, 56, interpolation='nearest')(x)
+    x = keras.layers.Add()([x, conv3])
+    x = FDDepthwiseBlock(x, 64, block_id=17)
+    x = keras.layers.experimental.preprocessing.Resizing(112, 112, interpolation='nearest')(x)
+    x = keras.layers.Add()([x, conv1])
+    x = FDDepthwiseBlock(x, 32, block_id=18)
+    x = keras.layers.experimental.preprocessing.Resizing(224, 224, interpolation='nearest')(x)
+
+    x = keras.layers.Conv2D(1, 1, padding='same')(x)
     x = keras.layers.BatchNormalization()(x)
-    x = keras.layers.ReLU()(x)
+    x = keras.layers.ReLU(6.)(x)
-    return keras.Model(input, x, name="fast_depth")
+    return keras.Model(inputs=input, outputs=x, name="fast_depth")
+
+
+if __name__ == '__main__':
+    make_fastdepth_functional().summary()

main.py

@@ -1,70 +1,6 @@
-import tensorflow as tf
-from tensorflow import keras
-
 import tensorflow_datasets as tfds
 
 
-class DecodeConv(keras.layers.Layer):
-    def __init__(self, out_filters, **kwargs):
-        super().__init__(**kwargs)
-        # Should be depthwise followed by batchnorm and relu.
-        self.depthwise = keras.layers.DepthwiseConv2D(5)
-        self.batch_norm = keras.layers.BatchNormalization()
-        self.relu = keras.layers.ReLU(6.)
-        self.pointwise = keras.layers.Conv2D(out_filters, 1)
-        self.pointwise_bn = keras.layers.BatchNormalization()
-        self.pointwise_rl = keras.layers.ReLU(6.)
-
-    def call(self, inputs, **kwargs):
-        inputs = self.depthwise(inputs, **kwargs)
-        inputs = self.batch_norm(inputs, **kwargs)
-        inputs = self.relu(inputs, **kwargs)
-        inputs = self.pointwise(inputs, **kwargs)
-        inputs = self.pointwise_bn(inputs, **kwargs)
-        return self.pointwise_rl(inputs, **kwargs)
-
-
-class FastDepth(keras.Model):
-    def get_config(self):
-        # TODO: What to put here?
-        pass
-
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        self.mobile_net = keras.applications.MobileNet(include_top=False)
-
-        # TODO: Try keras.layers.SeparableConv2D as well, should do the same thing if relu is used as activation
-        # It probably doesn't, since
-        self.decode_conv1 = DecodeConv(512)
-        self.decode_conv2 = DecodeConv(256)
-        self.decode_conv3 = DecodeConv(128)
-        self.decode_conv4 = DecodeConv(64)
-        self.decode_conv5 = DecodeConv(32)
-
-        self.final_pointwise = keras.layers.Conv2D(1, 1)
-        self.final_pointwise_bn = keras.layers.BatchNormalization()
-        self.final_pointwise_relu = keras.layers.ReLU()
-
-    def call(self, inputs, is_training=False, **kwargs):
-        # Go through mobilenet, then each decode layer, including skip connections using:
-        # keras.layers.Add()
-        inputs = self.mobile_net(inputs, is_training=is_training, **kwargs)
-
-        # FastDepth Additive Decoder
-        inputs = self.decode_conv1(inputs, is_training=is_training, **kwargs)
-        inputs = self.decode_conv2(inputs, is_training=is_training, **kwargs)
-        inputs = inputs + self.mobile_net.get_layer('conv_pw_5_relu').output
-        inputs = self.decode_conv3(inputs, is_training=is_training, **kwargs)
-        inputs = inputs + self.mobile_net.get_layer('conv_pw_3_relu').output
-        inputs = self.decode_conv4(inputs, is_training=is_training, **kwargs)
-        inputs = inputs + self.mobile_net.get_layer('conv_pw_1_relu').output
-        inputs = self.decode_conv5(inputs, is_training=is_training, **kwargs)
-
-        inputs = self.final_pointwise(inputs, is_training=is_training, **kwargs)
-        inputs = self.final_pointwise_bn(inputs, is_training=is_training, **kwargs)
-        return self.final_pointwise_relu(inputs, is_training=is_training, **kwargs)
-
-
 def load_nyu():
     builder = tfds.builder('nyu_depth_v2')
     builder.download_and_prepare(download_dir='../nyu')