Welcome to the final assignment of Week 3! You'll be building your own U-Net, a type of CNN designed for quick, precise image segmentation, and using it to predict a label for every single pixel in an image - in this case, an image from a self-driving car dataset.
This type of image classification is called semantic image segmentation. It's similar to object detection in that both ask the question: "What objects are in this image and where in the image are those objects located?," but where object detection labels objects with bounding boxes that may include pixels that aren't part of the object, semantic image segmentation allows you to predict a precise mask for each object in the image by labeling each pixel in the image with its corresponding class. The word “semantic” here refers to what's being shown, so for example the “Car” class is indicated below by the dark blue mask, and "Person" is indicated with a red mask:
As you might imagine, region-specific labeling is a pretty crucial consideration for self-driving cars, which require a pixel-perfect understanding of their environment so they can change lanes and avoid other cars, or any number of traffic obstacles that can put peoples' lives in danger.
By the time you finish this notebook, you'll be able to:
- Build your own U-Net
- Explain the difference between a regular CNN and a U-net
- Implement semantic image segmentation on the CARLA self-driving car dataset
- Apply sparse categorical crossentropy for pixelwise prediction
Onward, to this grand and glorious quest!
Run the cell below to import all the libraries you'll need:
import tensorflow as tf
import numpy as np
from tensorflow.keras.layers import Input
from tensorflow.keras.layers import Conv2D
from tensorflow.keras.layers import MaxPooling2D
from tensorflow.keras.layers import Dropout
from tensorflow.keras.layers import Conv2DTranspose
from tensorflow.keras.layers import concatenate
from test_utils import summary, comparatorimport os
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import imageio
import matplotlib.pyplot as plt
%matplotlib inline
path = ''
image_path = os.path.join(path, './data/CameraRGB/')
mask_path = os.path.join(path, './data/CameraMask/')
image_list = os.listdir(image_path)
mask_list = os.listdir(mask_path)
image_list = [image_path+i for i in image_list]
mask_list = [mask_path+i for i in mask_list]N = 2
img = imageio.imread(image_list[N])
mask = imageio.imread(mask_list[N])
#mask = np.array([max(mask[i, j]) for i in range(mask.shape[0]) for j in range(mask.shape[1])]).reshape(img.shape[0], img.shape[1])
fig, arr = plt.subplots(1, 2, figsize=(14, 10))
arr[0].imshow(img)
arr[0].set_title('Image')
arr[1].imshow(mask[:, :, 0])
arr[1].set_title('Segmentation')Text(0.5, 1.0, 'Segmentation')
image_list_ds = tf.data.Dataset.list_files(image_list, shuffle=False)
mask_list_ds = tf.data.Dataset.list_files(mask_list, shuffle=False)
for path in zip(image_list_ds.take(3), mask_list_ds.take(3)):
print(path)(<tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraRGB/000026.png'>, <tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraMask/000026.png'>)
(<tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraRGB/000027.png'>, <tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraMask/000027.png'>)
(<tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraRGB/000028.png'>, <tf.Tensor: shape=(), dtype=string, numpy=b'./data/CameraMask/000028.png'>)
image_filenames = tf.constant(image_list)
masks_filenames = tf.constant(mask_list)
dataset = tf.data.Dataset.from_tensor_slices((image_filenames, masks_filenames))
for image, mask in dataset.take(1):
print(image)
print(mask)tf.Tensor(b'./data/CameraRGB/002128.png', shape=(), dtype=string)
tf.Tensor(b'./data/CameraMask/002128.png', shape=(), dtype=string)
def process_path(image_path, mask_path):
img = tf.io.read_file(image_path)
img = tf.image.decode_png(img, channels=3)
img = tf.image.convert_image_dtype(img, tf.float32)
mask = tf.io.read_file(mask_path)
mask = tf.image.decode_png(mask, channels=3)
mask = tf.math.reduce_max(mask, axis=-1, keepdims=True)
return img, mask
def preprocess(image, mask):
input_image = tf.image.resize(image, (96, 128), method='nearest')
input_mask = tf.image.resize(mask, (96, 128), method='nearest')
input_image = input_image / 255.
return input_image, input_mask
image_ds = dataset.map(process_path)
processed_image_ds = image_ds.map(preprocess)U-Net, named for its U-shape, was originally created in 2015 for tumor detection, but in the years since has become a very popular choice for other semantic segmentation tasks.
U-Net builds on a previous architecture called the Fully Convolutional Network, or FCN, which replaces the dense layers found in a typical CNN with a transposed convolution layer that upsamples the feature map back to the size of the original input image, while preserving the spatial information. This is necessary because the dense layers destroy spatial information (the "where" of the image), which is an essential part of image segmentation tasks. An added bonus of using transpose convolutions is that the input size no longer needs to be fixed, as it does when dense layers are used.
Unfortunately, the final feature layer of the FCN suffers from information loss due to downsampling too much. It then becomes difficult to upsample after so much information has been lost, causing an output that looks rough.
U-Net improves on the FCN, using a somewhat similar design, but differing in some important ways. Instead of one transposed convolution at the end of the network, it uses a matching number of convolutions for downsampling the input image to a feature map, and transposed convolutions for upsampling those maps back up to the original input image size. It also adds skip connections, to retain information that would otherwise become lost during encoding. Skip connections send information to every upsampling layer in the decoder from the corresponding downsampling layer in the encoder, capturing finer information while also keeping computation low. These help prevent information loss, as well as model overfitting.
Contracting path (Encoder containing downsampling steps):
Images are first fed through several convolutional layers which reduce height and width, while growing the number of channels.
The contracting path follows a regular CNN architecture, with convolutional layers, their activations, and pooling layers to downsample the image and extract its features. In detail, it consists of the repeated application of two 3 x 3 unpadded convolutions, each followed by a rectified linear unit (ReLU) and a 2 x 2 max pooling operation with stride 2 for downsampling. At each downsampling step, the number of feature channels is doubled.
Crop function: This step crops the image from the contracting path and concatenates it to the current image on the expanding path to create a skip connection.
Expanding path (Decoder containing upsampling steps):
The expanding path performs the opposite operation of the contracting path, growing the image back to its original size, while shrinking the channels gradually.
In detail, each step in the expanding path upsamples the feature map, followed by a 2 x 2 convolution (the transposed convolution). This transposed convolution halves the number of feature channels, while growing the height and width of the image.
Next is a concatenation with the correspondingly cropped feature map from the contracting path, and two 3 x 3 convolutions, each followed by a ReLU. You need to perform cropping to handle the loss of border pixels in every convolution.
Final Feature Mapping Block: In the final layer, a 1x1 convolution is used to map each 64-component feature vector to the desired number of classes. The channel dimensions from the previous layer correspond to the number of filters used, so when you use 1x1 convolutions, you can transform that dimension by choosing an appropriate number of 1x1 filters. When this idea is applied to the last layer, you can reduce the channel dimensions to have one layer per class.
The U-Net network has 23 convolutional layers in total.
The encoder is a stack of various conv_blocks:
Each conv_block() is composed of 2 Conv2D layers with ReLU activations. We will apply Dropout, and MaxPooling2D to some conv_blocks, as you will verify in the following sections, specifically to the last two blocks of the downsampling.
The function will return two tensors:
next_layer: That will go into the next block.skip_connection: That will go into the corresponding decoding block.
Note: If max_pooling=True, the next_layer will be the output of the MaxPooling2D layer, but the skip_connection will be the output of the previously applied layer(Conv2D or Dropout, depending on the case). Else, both results will be identical.
Implement conv_block(...). Here are the instructions for each step in the conv_block, or contracting block:
- Add 2 Conv2D layers with
n_filtersfilters withkernel_sizeset to 3,kernel_initializerset to 'he_normal',paddingset to 'same' and 'relu' activation. - if
dropout_prob> 0, then add a Dropout layer with parameterdropout_prob - If
max_poolingis set to True, then add a MaxPooling2D layer with 2x2 pool size
# UNQ_C1
# GRADED FUNCTION: conv_block
def conv_block(inputs=None, n_filters=32, dropout_prob=0, max_pooling=True):
"""
Convolutional downsampling block
Arguments:
inputs -- Input tensor
n_filters -- Number of filters for the convolutional layers
dropout_prob -- Dropout probability
max_pooling -- Use MaxPooling2D to reduce the spatial dimensions of the output volume
Returns:
next_layer, skip_connection -- Next layer and skip connection outputs
"""
### START CODE HERE
conv = Conv2D(n_filters, # Number of filters
3, # Kernel size
activation='relu',
padding='same',
kernel_initializer= 'he_normal')(inputs)
conv = Conv2D(n_filters, # Number of filters
3, # Kernel size
activation='relu',
padding='same',
kernel_initializer='he_normal')(conv)
### END CODE HERE
# if dropout_prob > 0 add a dropout layer, with the variable dropout_prob as parameter
if dropout_prob > 0:
### START CODE HERE
conv = Dropout(dropout_prob)(conv)
### END CODE HERE
# if max_pooling is True add a MaxPooling2D with 2x2 pool_size
if max_pooling:
### START CODE HERE
next_layer = MaxPooling2D()(conv)
### END CODE HERE
else:
next_layer = conv
skip_connection = conv
return next_layer, skip_connectioninput_size=(96, 128, 3)
n_filters = 32
inputs = Input(input_size)
cblock1 = conv_block(inputs, n_filters * 1)
model1 = tf.keras.Model(inputs=inputs, outputs=cblock1)
output1 = [['InputLayer', [(None, 96, 128, 3)], 0],
['Conv2D', (None, 96, 128, 32), 896, 'same', 'relu', 'HeNormal'],
['Conv2D', (None, 96, 128, 32), 9248, 'same', 'relu', 'HeNormal'],
['MaxPooling2D', (None, 48, 64, 32), 0, (2, 2)]]
print('Block 1:')
for layer in summary(model1):
print(layer)
comparator(summary(model1), output1)
inputs = Input(input_size)
cblock1 = conv_block(inputs, n_filters * 32, dropout_prob=0.1, max_pooling=True)
model2 = tf.keras.Model(inputs=inputs, outputs=cblock1)
output2 = [['InputLayer', [(None, 96, 128, 3)], 0],
['Conv2D', (None, 96, 128, 1024), 28672, 'same', 'relu', 'HeNormal'],
['Conv2D', (None, 96, 128, 1024), 9438208, 'same', 'relu', 'HeNormal'],
['Dropout', (None, 96, 128, 1024), 0, 0.1],
['MaxPooling2D', (None, 48, 64, 1024), 0, (2, 2)]]
print('\nBlock 2:')
for layer in summary(model2):
print(layer)
comparator(summary(model2), output2)Block 1:
['InputLayer', [(None, 96, 128, 3)], 0]
['Conv2D', (None, 96, 128, 32), 896, 'same', 'relu', 'HeNormal']
['Conv2D', (None, 96, 128, 32), 9248, 'same', 'relu', 'HeNormal']
['MaxPooling2D', (None, 48, 64, 32), 0, (2, 2)]
�[32mAll tests passed!�[0m
Block 2:
['InputLayer', [(None, 96, 128, 3)], 0]
['Conv2D', (None, 96, 128, 1024), 28672, 'same', 'relu', 'HeNormal']
['Conv2D', (None, 96, 128, 1024), 9438208, 'same', 'relu', 'HeNormal']
['Dropout', (None, 96, 128, 1024), 0, 0.1]
['MaxPooling2D', (None, 48, 64, 1024), 0, (2, 2)]
�[32mAll tests passed!�[0m
The decoder, or upsampling block, upsamples the features back to the original image size. At each upsampling level, you'll take the output of the corresponding encoder block and concatenate it before feeding to the next decoder block.
There are two new components in the decoder: up and merge. These are the transpose convolution and the skip connections. In addition, there are two more convolutional layers set to the same parameters as in the encoder.
Here you'll encounter the Conv2DTranspose layer, which performs the inverse of the Conv2D layer. You can read more about it here.
Implement upsampling_block(...).
For the function upsampling_block:
- Takes the arguments
expansive_input(which is the input tensor from the previous layer) andcontractive_input(the input tensor from the previous skip layer) - The number of filters here is the same as in the downsampling block you completed previously
- Your
Conv2DTransposelayer will taken_filterswith shape (3,3) and a stride of (2,2), with padding set tosame. It's applied toexpansive_input, or the input tensor from the previous layer.
This block is also where you'll concatenate the outputs from the encoder blocks, creating skip connections.
- Concatenate your Conv2DTranspose layer output to the contractive input, with an
axisof 3. In general, you can concatenate the tensors in the order that you prefer. But for the grader, it is important that you use[up, contractive_input]
For the final component, set the parameters for two Conv2D layers to the same values that you set for the two Conv2D layers in the encoder (ReLU activation, He normal initializer, same padding).
# UNQ_C2
# GRADED FUNCTION: upsampling_block
def upsampling_block(expansive_input, contractive_input, n_filters=32):
"""
Convolutional upsampling block
Arguments:
expansive_input -- Input tensor from previous layer
contractive_input -- Input tensor from previous skip layer
n_filters -- Number of filters for the convolutional layers
Returns:
conv -- Tensor output
"""
### START CODE HERE
up = Conv2DTranspose(
n_filters, # number of filters
3, # Kernel size
strides=(2,2),
padding='same')(expansive_input)
# Merge the previous output and the contractive_input
merge = concatenate([up, contractive_input], axis=3)
conv = Conv2D(n_filters, # Number of filters
3, # Kernel size
activation='relu',
padding='same',
kernel_initializer='he_normal')(merge)
conv = Conv2D(n_filters, # Number of filters
3, # Kernel size
activation='relu',
padding='same',
kernel_initializer='he_normal')(conv)
### END CODE HERE
return convinput_size1=(12, 16, 256)
input_size2 = (24, 32, 128)
n_filters = 32
expansive_inputs = Input(input_size1)
contractive_inputs = Input(input_size2)
cblock1 = upsampling_block(expansive_inputs, contractive_inputs, n_filters * 1)
model1 = tf.keras.Model(inputs=[expansive_inputs, contractive_inputs], outputs=cblock1)
output1 = [['InputLayer', [(None, 12, 16, 256)], 0],
['Conv2DTranspose', (None, 24, 32, 32), 73760],
['InputLayer', [(None, 24, 32, 128)], 0],
['Concatenate', (None, 24, 32, 160), 0],
['Conv2D', (None, 24, 32, 32), 46112, 'same', 'relu', 'HeNormal'],
['Conv2D', (None, 24, 32, 32), 9248, 'same', 'relu', 'HeNormal']]
print('Block 1:')
for layer in summary(model1):
print(layer)
comparator(summary(model1), output1)Block 1:
['InputLayer', [(None, 12, 16, 256)], 0]
['Conv2DTranspose', (None, 24, 32, 32), 73760]
['InputLayer', [(None, 24, 32, 128)], 0]
['Concatenate', (None, 24, 32, 160), 0]
['Conv2D', (None, 24, 32, 32), 46112, 'same', 'relu', 'HeNormal']
['Conv2D', (None, 24, 32, 32), 9248, 'same', 'relu', 'HeNormal']
�[32mAll tests passed!�[0m
This is where you'll put it all together, by chaining the encoder, bottleneck, and decoder! You'll need to specify the number of output channels, which for this particular set would be 23. That's because there are 23 possible labels for each pixel in this self-driving car dataset.
For the function unet_model, specify the input shape, number of filters, and number of classes (23 in this case).
For the first half of the model:
- Begin with a conv block that takes the inputs of the model and the number of filters
- Then, chain the first output element of each block to the input of the next convolutional block
- Next, double the number of filters at each step
- Beginning with
conv_block4, adddropoutof 0.3 - For the final conv_block, set
dropoutto 0.3 again, and turn off max pooling
For the second half:
- Use cblock5 as expansive_input and cblock4 as contractive_input, with
n_filters* 8. This is your bottleneck layer. - Chain the output of the previous block as expansive_input and the corresponding contractive block output.
- Note that you must use the second element of the contractive block before the max pooling layer.
- At each step, use half the number of filters of the previous block
conv9is a Conv2D layer with ReLU activation, He normal initializer,samepadding- Finally,
conv10is a Conv2D that takes the number of classes as the filter, a kernel size of 1, and "same" padding. The output ofconv10is the output of your model.
# UNQ_C3
# GRADED FUNCTION: unet_model
def unet_model(input_size=(96, 128, 3), n_filters=32, n_classes=23):
"""
Unet model
Arguments:
input_size -- Input shape
n_filters -- Number of filters for the convolutional layers
n_classes -- Number of output classes
Returns:
model -- tf.keras.Model
"""
inputs = Input(input_size)
# Contracting Path (encoding)
# Add a conv_block with the inputs of the unet_ model and n_filters
### START CODE HERE
cblock1 = conv_block(inputs, n_filters)
# Chain the first element of the output of each block to be the input of the next conv_block.
# Double the number of filters at each new step
cblock2 = conv_block(cblock1[0], n_filters * 2)
cblock3 = conv_block(cblock2[0], n_filters * 4)
cblock4 = conv_block(cblock3[0], n_filters * 8, dropout_prob=0.3) # Include a dropout of 0.3 for this layer
# Include a dropout of 0.3 for this layer, and avoid the max_pooling layer
cblock5 = conv_block(cblock4[0], n_filters*16, dropout_prob=0.3, max_pooling=False)
### END CODE HERE
# Expanding Path (decoding)
# Add the first upsampling_block.
# Use the cblock5[0] as expansive_input and cblock4[1] as contractive_input and n_filters * 8
### START CODE HERE
ublock6 = upsampling_block(cblock5[0], cblock4[1], n_filters * 8)
# Chain the output of the previous block as expansive_input and the corresponding contractive block output.
# Note that you must use the second element of the contractive block i.e before the maxpooling layer.
# At each step, use half the number of filters of the previous block
ublock7 = upsampling_block(ublock6, cblock3[1], n_filters * 4)
ublock8 = upsampling_block(ublock7, cblock2[1], n_filters * 2)
ublock9 = upsampling_block(ublock8, cblock1[1], n_filters)
### END CODE HERE
conv9 = Conv2D(n_filters,
3,
activation='relu',
padding='same',
kernel_initializer='he_normal')(ublock9)
# Add a Conv2D layer with n_classes filter, kernel size of 1 and a 'same' padding
### START CODE HERE
conv10 = Conv2D(n_classes, 1, padding='same')(conv9)
### END CODE HERE
model = tf.keras.Model(inputs=inputs, outputs=conv10)
return modelimport outputs
img_height = 96
img_width = 128
num_channels = 3
unet = unet_model((img_height, img_width, num_channels))
comparator(summary(unet), outputs.unet_model_output)�[32mAll tests passed!�[0m
img_height = 96
img_width = 128
num_channels = 3
unet = unet_model((img_height, img_width, num_channels))unet.summary()Model: "functional_21"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_39 (InputLayer) [(None, 96, 128, 3)] 0
__________________________________________________________________________________________________
conv2d_415 (Conv2D) (None, 96, 128, 32) 896 input_39[0][0]
__________________________________________________________________________________________________
conv2d_416 (Conv2D) (None, 96, 128, 32) 9248 conv2d_415[0][0]
__________________________________________________________________________________________________
max_pooling2d_124 (MaxPooling2D (None, 48, 64, 32) 0 conv2d_416[0][0]
__________________________________________________________________________________________________
conv2d_417 (Conv2D) (None, 48, 64, 64) 18496 max_pooling2d_124[0][0]
__________________________________________________________________________________________________
conv2d_418 (Conv2D) (None, 48, 64, 64) 36928 conv2d_417[0][0]
__________________________________________________________________________________________________
max_pooling2d_125 (MaxPooling2D (None, 24, 32, 64) 0 conv2d_418[0][0]
__________________________________________________________________________________________________
conv2d_419 (Conv2D) (None, 24, 32, 128) 73856 max_pooling2d_125[0][0]
__________________________________________________________________________________________________
conv2d_420 (Conv2D) (None, 24, 32, 128) 147584 conv2d_419[0][0]
__________________________________________________________________________________________________
max_pooling2d_126 (MaxPooling2D (None, 12, 16, 128) 0 conv2d_420[0][0]
__________________________________________________________________________________________________
conv2d_421 (Conv2D) (None, 12, 16, 256) 295168 max_pooling2d_126[0][0]
__________________________________________________________________________________________________
conv2d_422 (Conv2D) (None, 12, 16, 256) 590080 conv2d_421[0][0]
__________________________________________________________________________________________________
dropout_56 (Dropout) (None, 12, 16, 256) 0 conv2d_422[0][0]
__________________________________________________________________________________________________
max_pooling2d_127 (MaxPooling2D (None, 6, 8, 256) 0 dropout_56[0][0]
__________________________________________________________________________________________________
conv2d_423 (Conv2D) (None, 6, 8, 512) 1180160 max_pooling2d_127[0][0]
__________________________________________________________________________________________________
conv2d_424 (Conv2D) (None, 6, 8, 512) 2359808 conv2d_423[0][0]
__________________________________________________________________________________________________
dropout_57 (Dropout) (None, 6, 8, 512) 0 conv2d_424[0][0]
__________________________________________________________________________________________________
conv2d_transpose_66 (Conv2DTran (None, 12, 16, 256) 1179904 dropout_57[0][0]
__________________________________________________________________________________________________
concatenate_49 (Concatenate) (None, 12, 16, 512) 0 conv2d_transpose_66[0][0]
dropout_56[0][0]
__________________________________________________________________________________________________
conv2d_425 (Conv2D) (None, 12, 16, 256) 1179904 concatenate_49[0][0]
__________________________________________________________________________________________________
conv2d_426 (Conv2D) (None, 12, 16, 256) 590080 conv2d_425[0][0]
__________________________________________________________________________________________________
conv2d_transpose_67 (Conv2DTran (None, 24, 32, 128) 295040 conv2d_426[0][0]
__________________________________________________________________________________________________
concatenate_50 (Concatenate) (None, 24, 32, 256) 0 conv2d_transpose_67[0][0]
conv2d_420[0][0]
__________________________________________________________________________________________________
conv2d_427 (Conv2D) (None, 24, 32, 128) 295040 concatenate_50[0][0]
__________________________________________________________________________________________________
conv2d_428 (Conv2D) (None, 24, 32, 128) 147584 conv2d_427[0][0]
__________________________________________________________________________________________________
conv2d_transpose_68 (Conv2DTran (None, 48, 64, 64) 73792 conv2d_428[0][0]
__________________________________________________________________________________________________
concatenate_51 (Concatenate) (None, 48, 64, 128) 0 conv2d_transpose_68[0][0]
conv2d_418[0][0]
__________________________________________________________________________________________________
conv2d_429 (Conv2D) (None, 48, 64, 64) 73792 concatenate_51[0][0]
__________________________________________________________________________________________________
conv2d_430 (Conv2D) (None, 48, 64, 64) 36928 conv2d_429[0][0]
__________________________________________________________________________________________________
conv2d_transpose_69 (Conv2DTran (None, 96, 128, 32) 18464 conv2d_430[0][0]
__________________________________________________________________________________________________
concatenate_52 (Concatenate) (None, 96, 128, 64) 0 conv2d_transpose_69[0][0]
conv2d_416[0][0]
__________________________________________________________________________________________________
conv2d_431 (Conv2D) (None, 96, 128, 32) 18464 concatenate_52[0][0]
__________________________________________________________________________________________________
conv2d_432 (Conv2D) (None, 96, 128, 32) 9248 conv2d_431[0][0]
__________________________________________________________________________________________________
conv2d_433 (Conv2D) (None, 96, 128, 32) 9248 conv2d_432[0][0]
__________________________________________________________________________________________________
conv2d_434 (Conv2D) (None, 96, 128, 23) 759 conv2d_433[0][0]
==================================================================================================
Total params: 8,640,471
Trainable params: 8,640,471
Non-trainable params: 0
__________________________________________________________________________________________________
In semantic segmentation, you need as many masks as you have object classes. In the dataset you're using, each pixel in every mask has been assigned a single integer probability that it belongs to a certain class, from 0 to num_classes-1. The correct class is the layer with the higher probability.
This is different from categorical crossentropy, where the labels should be one-hot encoded (just 0s and 1s). Here, you'll use sparse categorical crossentropy as your loss function, to perform pixel-wise multiclass prediction. Sparse categorical crossentropy is more efficient than other loss functions when you're dealing with lots of classes.
unet.compile(optimizer='adam',
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])Below, define a function that allows you to display both an input image, and its ground truth: the true mask. The true mask is what your trained model output is aiming to get as close to as possible.
def display(display_list):
plt.figure(figsize=(15, 15))
title = ['Input Image', 'True Mask', 'Predicted Mask']
for i in range(len(display_list)):
plt.subplot(1, len(display_list), i+1)
plt.title(title[i])
plt.imshow(tf.keras.preprocessing.image.array_to_img(display_list[i]))
plt.axis('off')
plt.show()for image, mask in image_ds.take(12):
sample_image, sample_mask = image, mask
print(mask.shape)
display([sample_image, sample_mask])(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
(480, 640, 1)
for image, mask in processed_image_ds.take(1):
sample_image, sample_mask = image, mask
print(mask.shape)
display([sample_image, sample_mask])(96, 128, 1)
EPOCHS = 40
VAL_SUBSPLITS = 5
BUFFER_SIZE = 500
BATCH_SIZE = 32
processed_image_ds.batch(BATCH_SIZE)
train_dataset = processed_image_ds.cache().shuffle(BUFFER_SIZE).batch(BATCH_SIZE)
print(processed_image_ds.element_spec)
model_history = unet.fit(train_dataset, epochs=EPOCHS)(TensorSpec(shape=(96, 128, 3), dtype=tf.float32, name=None), TensorSpec(shape=(96, 128, 1), dtype=tf.uint8, name=None))
Epoch 1/40
34/34 [==============================] - 21s 626ms/step - loss: 2.8165 - accuracy: 0.3202
Epoch 2/40
34/34 [==============================] - 3s 94ms/step - loss: 1.6785 - accuracy: 0.4142
Epoch 3/40
34/34 [==============================] - 3s 93ms/step - loss: 1.4179 - accuracy: 0.5422
Epoch 4/40
34/34 [==============================] - 3s 97ms/step - loss: 0.8158 - accuracy: 0.7731
Epoch 5/40
34/34 [==============================] - 3s 98ms/step - loss: 0.6694 - accuracy: 0.7973
Epoch 6/40
34/34 [==============================] - 3s 97ms/step - loss: 0.7657 - accuracy: 0.7721
Epoch 7/40
34/34 [==============================] - 3s 97ms/step - loss: 0.6299 - accuracy: 0.8031
Epoch 8/40
34/34 [==============================] - 3s 97ms/step - loss: 0.8957 - accuracy: 0.7608
Epoch 9/40
34/34 [==============================] - 3s 97ms/step - loss: 0.8036 - accuracy: 0.7669
Epoch 10/40
34/34 [==============================] - 3s 97ms/step - loss: 0.6417 - accuracy: 0.8050
Epoch 11/40
34/34 [==============================] - 3s 97ms/step - loss: 0.5888 - accuracy: 0.8125
Epoch 12/40
34/34 [==============================] - 3s 98ms/step - loss: 0.5763 - accuracy: 0.8163
Epoch 13/40
34/34 [==============================] - 3s 97ms/step - loss: 0.5286 - accuracy: 0.8293
Epoch 14/40
34/34 [==============================] - 3s 97ms/step - loss: 0.5801 - accuracy: 0.8170
Epoch 15/40
34/34 [==============================] - 3s 96ms/step - loss: 0.5125 - accuracy: 0.8337
Epoch 16/40
34/34 [==============================] - 3s 98ms/step - loss: 0.4665 - accuracy: 0.8486
Epoch 17/40
34/34 [==============================] - 3s 96ms/step - loss: 0.4359 - accuracy: 0.8613
Epoch 18/40
34/34 [==============================] - 3s 97ms/step - loss: 0.4301 - accuracy: 0.8656
Epoch 19/40
34/34 [==============================] - 3s 97ms/step - loss: 0.3775 - accuracy: 0.8820
Epoch 20/40
34/34 [==============================] - 3s 97ms/step - loss: 0.3559 - accuracy: 0.8907
Epoch 21/40
34/34 [==============================] - 3s 92ms/step - loss: 0.3253 - accuracy: 0.9007
Epoch 22/40
34/34 [==============================] - 3s 92ms/step - loss: 0.2977 - accuracy: 0.9094
Epoch 23/40
34/34 [==============================] - 3s 92ms/step - loss: 0.3055 - accuracy: 0.9072
Epoch 24/40
34/34 [==============================] - 3s 91ms/step - loss: 0.2720 - accuracy: 0.9173
Epoch 25/40
34/34 [==============================] - 3s 92ms/step - loss: 0.2554 - accuracy: 0.9228
Epoch 26/40
34/34 [==============================] - 3s 92ms/step - loss: 0.2512 - accuracy: 0.9231
Epoch 27/40
34/34 [==============================] - 3s 93ms/step - loss: 0.2429 - accuracy: 0.9262
Epoch 28/40
34/34 [==============================] - 3s 92ms/step - loss: 0.2376 - accuracy: 0.9275
Epoch 29/40
34/34 [==============================] - 3s 93ms/step - loss: 0.2225 - accuracy: 0.9321
Epoch 30/40
34/34 [==============================] - 3s 85ms/step - loss: 0.2163 - accuracy: 0.9340
Epoch 31/40
34/34 [==============================] - 3s 86ms/step - loss: 0.2056 - accuracy: 0.9370
Epoch 32/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1999 - accuracy: 0.9386
Epoch 33/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1996 - accuracy: 0.9384
Epoch 34/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1916 - accuracy: 0.9409
Epoch 35/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1835 - accuracy: 0.9433
Epoch 36/40
34/34 [==============================] - 3s 91ms/step - loss: 0.1758 - accuracy: 0.9456
Epoch 37/40
34/34 [==============================] - 3s 91ms/step - loss: 0.1887 - accuracy: 0.9414
Epoch 38/40
34/34 [==============================] - 3s 91ms/step - loss: 0.1740 - accuracy: 0.9457
Epoch 39/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1705 - accuracy: 0.9465
Epoch 40/40
34/34 [==============================] - 3s 92ms/step - loss: 0.1621 - accuracy: 0.9493
Now, define a function that uses tf.argmax in the axis of the number of classes to return the index with the largest value and merge the prediction into a single image:
def create_mask(pred_mask):
pred_mask = tf.argmax(pred_mask, axis=-1)
pred_mask = pred_mask[..., tf.newaxis]
return pred_mask[0]Let's see how your model did!
plt.plot(model_history.history["accuracy"])[<matplotlib.lines.Line2D at 0x7fc2a0be2a20>]
Next, check your predicted masks against the true mask and the original input image:
def show_predictions(dataset=None, num=1):
"""
Displays the first image of each of the num batches
"""
if dataset:
for image, mask in dataset.take(num):
pred_mask = unet.predict(image)
display([image[0], mask[0], create_mask(pred_mask)])
else:
display([sample_image, sample_mask,
create_mask(unet.predict(sample_image[tf.newaxis, ...]))])show_predictions(train_dataset, 6)
With 40 epochs you get amazing results!
You've come to the end of this assignment. Awesome work creating a state-of-the art model for semantic image segmentation! This is a very important task for self-driving cars to get right. Elon Musk will surely be knocking down your door at any moment. ;)
What you should remember:
- Semantic image segmentation predicts a label for every single pixel in an image
- U-Net uses an equal number of convolutional blocks and transposed convolutions for downsampling and upsampling
- Skip connections are used to prevent border pixel information loss and overfitting in U-Net