Practical Convolutional Neural Network Models: Enhance deep learning skills by building intelligent ConvNet models using Keras
Mohit Sewak, Md. Rezaul Karim, Pradeep Pujari
- 出版商: Packt Publishing
- 出版日期: 2018-02-26
- 售價: $1,160
- 貴賓價: 9.5 折 $1,102
- 語言: 英文
- 頁數: 218
- 裝訂: Paperback
- ISBN: 1788392302
- ISBN-13: 9781788392303
實用捲積神經網絡：運用 Python 實現高級深度學習模型 (Practical Convolutional Neural Network Models: Enhance deep learning skills by building intelligent ConvNet models using Keras) (簡中版)
One stop guide to practice ConvNets models from most common to recent advances in artificial intelligence field
- Fast-paced guide with use cases and real-world examples to get well versed with CNN techniques
- Learn best practices to get the most out of the book to implement CNN models on image recognition, object classification, transfer learning, GAN and more
- Develop your skills by in-depth understanding of advanced CNN architectures such as AlexNet, VGG, GoogLeNet and more and apply them to real-world research field
Convolutional Neural Network (CNN) is revolutionizing several application domains such as visual recognition systems, self-driving cars, speech recognition and more. These advances create unprecedented opportunities and challenges to build and deploy large-scale ConvNet applications. This book aims to take you through the building blocks of CNN's, best practices for implementing CNN models and how they can be applied to solve complex machine learning problems.
This book starts with an overview of deep neural networks with the example of image classification and walks you through building your first CNN for human face detector. As you progress further you'll come across practical illustrations of CNN internals with interesting examples to understand different optimization and visualization techniques to build a robust model. While explaining CNN architecture, this book covers case studies of most common and award-winning CNN architectures.
Furthermore, this book examines how knowledge transfer can be achieved to train a CNN model that does need a lot of data. You will be introduced to the concept of transfer learning that helps to improve the performance of a CNN model besides training data needs. Towards the end, this book also touches upon the subject on the attention-based CNN with an example of visual question answering application. Finally, this book covers details of generative models and a novel application - getting started with generating your own hand-written MNIST digits.
By the end of this book, you will be all ready to implement CNN models in your work or projects by working with extreme datasets.
What you will learn
- From CNN basics building blocks to advanced concepts understand the practical areas they can be applied to.
- Build a simple image classifier CNN model to understand how different components interact with each other.
- Learn CNN Model Optimization and Visualization techniques
- Implement award-winning CNN Architectures like AlexNet, VGG, GoogLeNet, ResNet etc
- Practical approach to use pre-trained models and examples to understand transfer learning methodology.
- Understand the difference between GAN generators and discriminators.
- Step into the world of Artificial intelligence with adversarial training and applications of GANs
Who This Book Is For
This book is for data scientists, machine learning practitioners, deep learning and AI enthusiasts who want to move one step further in building convolutional neural network models. Get your hands on extreme datasets and different CNN architectures to build efficient and smart ConvNet models. Basic knowledge of machine learning and CNN is expected.