Practical convolutional neural networks : implement advanced deep learning models using Python

목차 (Table of Contents)

• CONTENTS
• Preface = 1
• Chapter 1 Deep Neural Networks - Overview = 7
• Building blocks of a neural network = 8
• Introduction to TensorFlow = 9
• Installing TensorFlow = 10
• Introduction to the MNIST dataset = 16
• The simplest artificial neural network = 17
• Keras deep learning library overview = 21
• Layers in the Keras model = 22
• Handwritten number recognition with Keras and MNIST = 23
• Retrieving training and test data = 24
• Visualizing the training data = 25
• Building the network = 26
• Training the network = 27
• Testing = 27
• Understanding backpropagation = 27
• Summary = 30
• Chapter 2 Introduction to Convolutional Neural Networks = 31
• History of CNNs = 32
• Convolutional neural networks = 33
• How do computers interpret images? = 35
• Code for visualizing an image = 35
• Dropout = 38
• Input layer = 38
• Convolutional layer = 39
• Pooling layer = 41
• Practical example - image classification = 42
• Image augmentation = 44
• Summary = 46
• Chapter 3 Build Your First CNN and Performance Optimization = 47
• CNN architectures and drawbacks of DNNs = 48
• Convolutional operations = 51
• Pooling, stride, and padding operations = 53
• Convolution and pooling operations in TensorFlow = 55
• Applying pooling operations in TensorFlow = 56
• Convolution operations in TensorFlow = 58
• Training a CNN = 60
• Weight and bias initialization = 60
• Regularization = 61
• Activation functions = 62
• Building, training, and evaluating our first CNN = 63
• Dataset description = 63
• Model performance optimization = 80
• Number of hidden layers = 81
• Number of neurons per hidden layer = 81
• Batch normalization = 82
• Advanced regularization and avoiding overfitting = 85
• Which optimizer to use? = 87
• Memory tuning = 88
• Appropriate layer placement = 88
• Building the second CNN by putting everything together = 89
• Summary = 98
• Chapter 4 Popular CNN Model Architectures = 99
• Introduction to ImageNet = 99
• LeNet = 101
• AlexNet architecture = 101
• Traffic sign classifiers using AlexNet = 103
• VGGNet architecture = 103
• VGG16 image classification code example = 105
• GoogLeNet architecture = 106
• Architecture insights = 106
• Inception module = 108
• ResNet architecture = 108
• Summary = 110
• Chapter 5 Transfer Learning = 111
• Feature extraction approach = 111
• Target dataset is small and is similar to the original training dataset = 113
• Target dataset is small but different from the original training dataset = 114
• Target dataset is large and similar to the original training dataset = 116
• Target dataset is large and different from the original training dataset = 117
• Transfer learning example = 118
• Multi-task learning = 120
• Summary = 121
• Chapter 6 Autoencoders for CNN = 123
• Introducing to autoencoders = 123
• Convolutional autoencoder = 124
• Applications = 125
• An example of compression = 125
• Summary = 126
• Chapter 7 Object Detection and Instance Segmentation with CNN = 129
• The differences between object detection and image classification = 130
• Why is object detection much more challenging than image classification? = 132
• Traditional, nonCNN approaches to object detection = 134
• Haar features, cascading classifiers, and the Viola-Jones algorithm = 134
• R-CNN - Regions with CNN features = 138
• Fast R-CNN - fast region-based CNN = 140
• Faster R-CNN - faster region proposal network-based CNN = 142
• Mask R-CNN - Instance segmentation with CNN = 145
• Instance segmentation in code = 148
• Creating the environment = 148
• Preparing the COCO dataset folder structure = 149
• Running the pre-trained model on the COCO dataset = 150
• References = 150
• Summary = 152
• Chapter 8 GAN : Generating New Images with CNN = 153
• Pix2pix - Image-to-Image translation GAN = 154
• CycleGAN = 155
• Training a GAN model = 157
• GAN - code example = 157
• Calculating loss = 160
• Semi-supervised learning and GAN = 163
• Feature matching = 164
• Semi-supervised classification using a GAN example = 164
• Deep convolutional GAN = 170
• Summary = 172
• Chapter 9 Attention Mechanism for CNN and Visual Models = 173
• Attention mechanism for image captioning = 176
• Types of Attention = 180
• Hard Attention = 181
• Soft Attention = 182
• Using attention to improve visual models = 183
• Reasons for sub-optimal performance of visual CNN models = 183
• Recurrent models of visual attention = 188
• References = 195
• Summary = 196
• Other Books You May Enjoy = 197
• Index = 201