def up_layer(x1, x2, in_size1, in_size2, out_size, i):
# Up 1
W1 = utils.weight_variable([2, 2, 2, in_size2, in_size1],
name="W_u_" + str(i) + "_1")
b1 = utils.bias_variable([in_size2], name="b_u_" + str(i) + "_1")
deco1 = utils.conv3d_transpose_strided(x1,
W1,
b1,
output_shape=tf.shape(x2))
relu1 = tf.nn.relu(deco1, name="relu_d_" + str(i) + "_1")
# Concat
conc1 = tf.concat([relu1, x2],
-1) # concat along the channels dimension
# Conv1
W2 = utils.weight_variable([3, 3, 3, in_size2 * 2, out_size],
name="W_u_" + str(i) + "_2")
b2 = utils.bias_variable([out_size], name="b_u_" + str(i) + "_2")
conv1 = utils.conv3d_basic(conc1, W2, b2)
relu2 = tf.nn.relu(conv1, name="relu_u_" + str(i) + "_2")
relu2 = tf.nn.dropout(relu2, keep_prob=keep_prob)
# Conv2
W3 = utils.weight_variable([3, 3, 3, out_size, out_size],
name="W_u_" + str(i) + "_3")
b3 = utils.bias_variable([out_size], name="b_u_" + str(i) + "_3")
conv3 = utils.conv3d_basic(relu2, W3, b3)
relu3 = tf.nn.relu(conv3, name="relu_u_" + str(i) + "_3")
relu3 = tf.nn.dropout(relu3, keep_prob=keep_prob)
return relu3
def down_layer(x, in_size, out_size, i):
# Down 1
W1 = utils.weight_variable([3, 3, 3, in_size, out_size // 2],
name="W_d_" + str(i) + "_1")
b1 = utils.bias_variable([out_size // 2],
name="b_d_" + str(i) + "_1")
conv1 = utils.conv3d_basic(x, W1, b1)
relu1 = tf.nn.relu(conv1, name="relu_d_" + str(i) + "_1")
relu1 = tf.nn.dropout(relu1, keep_prob=keep_prob)
# Down 2
W2 = utils.weight_variable([3, 3, 3, out_size // 2, out_size],
name="W_d_" + str(i) + "_2")
b2 = utils.bias_variable([out_size], name="b_d_" + str(i) + "_2")
conv2 = utils.conv3d_basic(relu1, W2, b2)
relu2 = tf.nn.relu(conv2, name="relu_d_" + str(i) + "_2")
relu2 = tf.nn.dropout(relu2, keep_prob=keep_prob)
# Pool 1
pool = utils.max_pool_2x2x2(relu2)
return relu2, pool
def inference(image, keep_prob):
"""
Semantic segmentation network definition
:param image: input image. Should have values in range 0-255
:param keep_prob:
:return:
"""
#using architecture described in https://arxiv.org/pdf/1704.06382.pdf
#TODO: Look up preprocessing step
processed_image = image / 100
with tf.variable_scope("inference"):
def down_layer(x, in_size, out_size, i):
# Down 1
W1 = utils.weight_variable([3, 3, 3, in_size, out_size // 2],
name="W_d_" + str(i) + "_1")
b1 = utils.bias_variable([out_size // 2],
name="b_d_" + str(i) + "_1")
conv1 = utils.conv3d_basic(x, W1, b1)
relu1 = tf.nn.relu(conv1, name="relu_d_" + str(i) + "_1")
relu1 = tf.nn.dropout(relu1, keep_prob=keep_prob)
# Down 2
W2 = utils.weight_variable([3, 3, 3, out_size // 2, out_size],
name="W_d_" + str(i) + "_2")
b2 = utils.bias_variable([out_size], name="b_d_" + str(i) + "_2")
conv2 = utils.conv3d_basic(relu1, W2, b2)
relu2 = tf.nn.relu(conv2, name="relu_d_" + str(i) + "_2")
relu2 = tf.nn.dropout(relu2, keep_prob=keep_prob)
# Pool 1
pool = utils.max_pool_2x2x2(relu2)
return relu2, pool
# Apply 4 down layes of increasing sizes
d1, p1 = down_layer(processed_image, 1, 64, 1)
d2, p2 = down_layer(p1, 64, 128, 2)
d3, p3 = down_layer(p2, 128, 256, 3)
d4, p4 = down_layer(p3, 256, 512, 4)
def up_layer(x1, x2, in_size1, in_size2, out_size, i):
# Up 1
W1 = utils.weight_variable([2, 2, 2, in_size2, in_size1],
name="W_u_" + str(i) + "_1")
b1 = utils.bias_variable([in_size2], name="b_u_" + str(i) + "_1")
deco1 = utils.conv3d_transpose_strided(x1,
W1,
b1,
output_shape=tf.shape(x2))
relu1 = tf.nn.relu(deco1, name="relu_d_" + str(i) + "_1")
# Concat
conc1 = tf.concat([relu1, x2],
-1) # concat along the channels dimension
# Conv1
W2 = utils.weight_variable([3, 3, 3, in_size2 * 2, out_size],
name="W_u_" + str(i) + "_2")
b2 = utils.bias_variable([out_size], name="b_u_" + str(i) + "_2")
conv1 = utils.conv3d_basic(conc1, W2, b2)
relu2 = tf.nn.relu(conv1, name="relu_u_" + str(i) + "_2")
relu2 = tf.nn.dropout(relu2, keep_prob=keep_prob)
# Conv2
W3 = utils.weight_variable([3, 3, 3, out_size, out_size],
name="W_u_" + str(i) + "_3")
b3 = utils.bias_variable([out_size], name="b_u_" + str(i) + "_3")
conv3 = utils.conv3d_basic(relu2, W3, b3)
relu3 = tf.nn.relu(conv3, name="relu_u_" + str(i) + "_3")
relu3 = tf.nn.dropout(relu3, keep_prob=keep_prob)
return relu3
#Apply 3 Up layers with skip connections
u3 = up_layer(d4, d3, 512, 256, 256, 3)
u2 = up_layer(u3, d2, 256, 128, 128, 2)
u1 = up_layer(u2, d1, 128, 64, 64, 1)
#Apply a final Conv layer
W = utils.weight_variable([1, 1, 1, 64, NUM_OF_CLASSES], name="W_o")
b = utils.bias_variable([NUM_OF_CLASSES], name="b_o")
conv = utils.conv3d_basic(u1, W, b)
annotation_pred = tf.argmax(conv, dimension=4, name="prediction")
return tf.expand_dims(annotation_pred, dim=4), conv
