def test_init_UNet():
"""
Testing init of UNet as expected
"""
local_test = u.UNet(
image_size=[1, 2, 3],
out_channels=3,
num_channel_initial=3,
depth=5,
out_kernel_initializer="he_normal",
out_activation="softmax",
)
# Asserting num channels initial is the same, Pass
assert local_test._num_channel_initial == 3
# Asserting depth is the same, Pass
assert local_test._depth == 5
# Assert downsample blocks type is correct, Pass
assert all(
isinstance(item, type(layer.DownSampleResnetBlock(12)))
for item in local_test._downsample_blocks
)
# Assert number of downsample blocks is correct (== depth), Pass
assert len(local_test._downsample_blocks) == 5
# Assert bottom_conv3d type is correct, Pass
assert isinstance(local_test._bottom_conv3d, type(layer.Conv3dBlock(5)))
# Assert bottom res3d type is correct, Pass
assert isinstance(local_test._bottom_res3d, type(layer.Residual3dBlock(5)))
# Assert upsample blocks type is correct, Pass
assert all(
isinstance(item, type(layer.UpSampleResnetBlock(12)))
for item in local_test._upsample_blocks
)
# Assert number of upsample blocks is correct (== depth), Pass
assert len(local_test._upsample_blocks) == 5
# Assert output_conv3d is correct type, Pass
assert isinstance(
local_test._output_conv3d, type(layer.Conv3dWithResize([1, 2, 3], 3))
)
def test_unet_return():
"""
Testing that build_backbone func returns an object
of type UNet form backbone module when initialised
with the associated UNet config.
"""
out = util.build_backbone(
image_size=(1, 2, 3),
out_channels=1,
model_config={
"backbone": "unet",
"unet": {
"num_channel_initial": 4,
"depth": 4
},
},
method_name="ddf",
)
assert isinstance(
out, type(u_net.UNet([1, 2, 3], 4, 4, 4, "he_normal", "sigmoid")))
def test_call_unet(self, image_size, depth):
out = 3
# initialising UNet instance
network = u.UNet(
image_size=image_size,
out_channels=3,
num_channel_initial=2,
depth=depth,
out_kernel_initializer="he_normal",
out_activation="softmax",
)
# pass an input of all zeros
inputs = tf.constant(
np.zeros(
(5, image_size[0], image_size[1], image_size[2], out), dtype=np.float32
)
)
# get outputs by calling
output = network.call(inputs)
# expected shape is (5, 1, 2, 3)
assert all(x == y for x, y in zip(inputs.shape, output.shape))
def test_call_UNet():
"""
Asserting that output shape of UNet call method
is correct.
"""
out = 3
im_size = [1, 2, 3]
# Initialising LocalNet instance
global_test = u.UNet(
image_size=im_size,
out_channels=out,
num_channel_initial=3,
depth=6,
out_kernel_initializer="glorot_uniform",
out_activation="sigmoid",
)
# Pass an input of all zeros
inputs = np.zeros((5, im_size[0], im_size[1], im_size[2], 3))
# Get outputs by calling
output = global_test.call(inputs)
# Expected shape is (5, 1, 2, 3)
assert all(x == y for x, y in zip(inputs.shape, output.shape))
def test_init(self, image_size, depth):
network = u.UNet(
image_size=image_size,
out_channels=3,
num_channel_initial=2,
depth=depth,
out_kernel_initializer="he_normal",
out_activation="softmax",
)
# asserting num channels initial is the same, Pass
assert network._num_channel_initial == 2
# asserting depth is the same, Pass
assert network._depth == depth
# assert downsample blocks type is correct, Pass
assert all(
isinstance(item, layer.DownSampleResnetBlock)
for item in network._downsample_blocks
)
# assert number of downsample blocks is correct (== depth), Pass
assert len(network._downsample_blocks) == depth
# assert bottom_conv3d type is correct, Pass
assert isinstance(network._bottom_conv3d, layer.Conv3dBlock)
# assert bottom res3d type is correct, Pass
assert isinstance(network._bottom_res3d, layer.Residual3dBlock)
# assert upsample blocks type is correct, Pass
assert all(
isinstance(item, layer.UpSampleResnetBlock)
for item in network._upsample_blocks
)
# assert number of upsample blocks is correct (== depth), Pass
assert len(network._upsample_blocks) == depth
# assert output_conv3d is correct type, Pass
assert isinstance(network._output_conv3d, layer.Conv3dWithResize)
