| Description |
1 online resource (xxiii, 282 pages) : illustrations (some color) |
| Contents |
880-01 Preface; Books Website; Contents; Acronyms; List of Figures; 1 Traffic Sign Detection and Recognition; 1.1 Introduction; 1.2 Challenges; 1.3 Previous Work; 1.3.1 Template Matching; 1.3.2 Hand-Crafted Features; 1.3.3 Feature Learning; 1.3.4 ConvNets; 1.4 Summary; 2 Pattern Classification; 2.1 Formulation; 2.1.1 K-Nearest Neighbor; 2.2 Linear Classifier; 2.2.1 Training a Linear Classifier; 2.2.2 Hinge Loss; 2.2.3 Logistic Regression; 2.2.4 Comparing Loss Function; 2.3 Multiclass Classification; 2.3.1 One Versus One; 2.3.2 One Versus Rest; 2.3.3 Multiclass Hinge Loss |
|
880-01/(S 2.3.4 Multinomial Logistic Function2.4 Feature Extraction; 2.5 Learning Φ(x); 2.6 Artificial Neural Networks; 2.6.1 Backpropagation; 2.6.2 Activation Functions; 2.6.3 Role of Bias; 2.6.4 Initialization; 2.6.5 How to Apply on Images; 2.7 Summary; 2.8 Exercises; 3 Convolutional Neural Networks; 3.1 Deriving Convolution from a Fully Connected Layer; 3.1.1 Role of Convolution; 3.1.2 Backpropagation of Convolution Layers; 3.1.3 Stride in Convolution; 3.2 Pooling; 3.2.1 Backpropagation in Pooling Layer; 3.3 LeNet; 3.4 AlexNet; 3.5 Designing a ConvNet; 3.5.1 ConvNet Architecture |
|
3.5.2 Software Libraries3.5.3 Evaluating a ConvNet; 3.6 Training a ConvNet; 3.6.1 Loss Function; 3.6.2 Initialization; 3.6.3 Regularization; 3.6.4 Learning Rate Annealing; 3.7 Analyzing Quantitative Results; 3.8 Other Types of Layers; 3.8.1 Local Response Normalization; 3.8.2 Spatial Pyramid Pooling; 3.8.3 Mixed Pooling; 3.8.4 Batch Normalization; 3.9 Summary; 3.10 Exercises; 4 Caffe Library; 4.1 Introduction; 4.2 Installing Caffe; 4.3 Designing Using Text Files; 4.3.1 Providing Data; 4.3.2 Convolution Layers; 4.3.3 Initializing Parameters; 4.3.4 Activation Layer; 4.3.5 Pooling Layer |
|
4.3.6 Fully Connected Layer4.3.7 Dropout Layer; 4.3.8 Classification and Loss Layers; 4.4 Training a Network; 4.5 Designing in Python; 4.6 Drawing Architecture of Network; 4.7 Training Using Python; 4.8 Evaluating Using Python; 4.9 Save and Restore Networks; 4.10 Python Layer in Caffe; 4.11 Summary; 4.12 Exercises; 5 Classification of Traffic Signs; 5.1 Introduction; 5.2 Related Work; 5.2.1 Template Matching; 5.2.2 Hand-Crafted Features; 5.2.3 Sparse Coding; 5.2.4 Discussion; 5.2.5 ConvNets; 5.3 Preparing Dataset; 5.3.1 Splitting Data; 5.3.2 Augmenting Dataset |
|
5.3.3 Static Versus One-the-Fly Augmenting5.3.4 Imbalanced Dataset; 5.3.5 Preparing the GTSRB Dataset; 5.4 Analyzing Training/Validation Curves; 5.5 ConvNets for Classification of Traffic Signs; 5.6 Ensemble of ConvNets; 5.6.1 Combining Models; 5.6.2 Training Different Models; 5.6.3 Creating Ensemble; 5.7 Evaluating Networks; 5.7.1 Misclassified Images; 5.7.2 Cross-Dataset Analysis and Transfer Learning; 5.7.3 Stability of ConvNet; 5.7.4 Analyzing by Visualization; 5.8 Analyzing by Visualizing; 5.8.1 Visualizing Sensitivity; 5.8.2 Visualizing the Minimum Perception |
| Summary |
This must-read text/reference introduces the fundamental concepts of convolutional neural networks (ConvNets), offering practical guidance on using libraries to implement ConvNets in applications of traffic sign detection and classification. The work presents techniques for optimizing the computational efficiency of ConvNets, as well as visualization techniques to better understand the underlying processes. The proposed models are also thoroughly evaluated from different perspectives, using exploratory and quantitative analysis. Topics and features: Explains the fundamental concepts behind training linear classifiers and feature learning Discusses the wide range of loss functions for training binary and multi-class classifiers Illustrates how to derive ConvNets from fully connected neural networks, and reviews different techniques for evaluating neural networks Presents a practical library for implementing ConvNets, explaining how to use a Python interface for the library to create and assess neural networks Describes two real-world examples of the detection and classification of traffic signs using deep learning methods Examines a range of varied techniques for visualizing neural networks, using a Python interface Provides self-study exercises at the end of each chapter, in addition to a helpful glossary, with relevant Python scripts supplied at an associated website This self-contained guide will benefit those who seek to both understand the theory behind deep learning, and to gain hands-on experience in implementing ConvNets in practice. As no prior background knowledge in the field is required to follow the material, the book is ideal for all students of computer vision and machine learning, and will also be of great interest to practitioners working on autonomous cars and advanced driver assistance systems |
| Bibliography |
Includes bibliographical references and index |
| Notes |
Online resource; title from PDF title page (SpringerLink, viewed June 1, 2017) |
| Subject |
Neural networks (Computer science)
|
|
Traffic signs and signals.
|
|
Computer vision.
|
|
Neural Networks, Computer
|
|
Information retrieval.
|
|
Computer networking & communications.
|
|
Imaging systems & technology.
|
|
Natural language & machine translation.
|
|
Automotive technology & trades.
|
|
Pattern recognition.
|
|
COMPUTERS -- General.
|
|
Computer vision
|
|
Neural networks (Computer science)
|
|
Traffic signs and signals
|
| Form |
Electronic book
|
| Author |
Heravi, Elnaz Jahani, author
|
| ISBN |
9783319575506 |
|
3319575503 |
|
