Advanced Deep Learning with TensorFlow 2 and Keras: Apply DL, GANs, VAEs, deep RL, unsupervised learning, object detection and segmentation, and more, 2nd Edition

by Rowel Atienza

Updated and revised second edition of the bestselling guide to advanced deep learning with TensorFlow 2 and Keras Advanced Deep Learning with TensorFlow 2 and Keras, Second Edition is a completely updated edition of the bestselling guide to the advanced deep learning techniques available today. Revised for TensorFlow 2.x, this edition introduces you to the practical side of deep learning with new chapters on unsupervised learning using mutual information, object detection (SSD), and semantic segmentation (FCN and PSPNet), further allowing you to create your own cutting-edge AI projects. Using Keras as an open-source deep learning library, the book features hands-on projects that show you how to create more effective AI with the most up-to-date techniques. Starting with an overview of multi-layer perceptrons (MLPs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs), the book then introduces more cutting-edge techniques as you explore deep neural network architectures, including ResNet and DenseNet, and how to create autoencoders. You will then learn about GANs, and how they can unlock new levels of AI performance. Next, you'll discover how a variational autoencoder (VAE) is implemented, and how GANs and VAEs have the generative power to synthesize data that can be extremely convincing to humans. You'll also learn to implement DRL such as Deep Q-Learning and Policy Gradient Methods, which are critical to many modern results in AI. This is not an introductory book, so fluency with Python is required. The reader should also be familiar with some machine learning approaches, and practical experience with DL will also be helpful. Knowledge of Keras or TensorFlow 2.0 is not required but is recommended.

ebook

Published February 28, 2020

Reviews

[Walter Ullon · Rating 5 out of 5]

Unlike other authors that release new editions yearly with few revisions, Atienza has packed more than 40% worth of content versus the 1st edition. This is a first for me.

The 1st edition is just about identical to the 2nd up to chapter 10, save for some stylistic changes, but half of chapter 10, as well as chapters 11, 12, 13, are brand new.

Everything that made the 1st edition great is till here: great exposition of the fundamentals, the math for those that'd like to dig a bit deeper, the great references, as well as clean, understandable, working code.

This book is highly recommended as a reference for those looking to move into deeper waters in the production of ML models for use cases that would strain conventional Neural Networks.

If I could, however, make a suggestion to all authors of ML books, it would be to PLEASE move away from MNIST, and the CIFAR datasets. I know these are the standard benchmarks in the field but you could just as easily cite this in your repositories and use the real estate in your books to explore other use cases such as anomaly detection in categorical data, time series, etc. This would make your books much more valuable in the hands of industry practitioners not looking to work strictly with images. So let's stop beating those horses yes?

[Benjamin · Rating 3 out of 5]

Honestly, a slog. I wish they spent more time on supervised style transfer