Book Image

The Deep Learning Workshop

By : Mirza Rahim Baig, Thomas V. Joseph, Nipun Sadvilkar, Mohan Kumar Silaparasetty, Anthony So
Book Image

The Deep Learning Workshop

By: Mirza Rahim Baig, Thomas V. Joseph, Nipun Sadvilkar, Mohan Kumar Silaparasetty, Anthony So

Overview of this book

Are you fascinated by how deep learning powers intelligent applications such as self-driving cars, virtual assistants, facial recognition devices, and chatbots to process data and solve complex problems? Whether you are familiar with machine learning or are new to this domain, The Deep Learning Workshop will make it easy for you to understand deep learning with the help of interesting examples and exercises throughout. The book starts by highlighting the relationship between deep learning, machine learning, and artificial intelligence and helps you get comfortable with the TensorFlow 2.0 programming structure using hands-on exercises. You’ll understand neural networks, the structure of a perceptron, and how to use TensorFlow to create and train models. The book will then let you explore the fundamentals of computer vision by performing image recognition exercises with convolutional neural networks (CNNs) using Keras. As you advance, you’ll be able to make your model more powerful by implementing text embedding and sequencing the data using popular deep learning solutions. Finally, you’ll get to grips with bidirectional recurrent neural networks (RNNs) and build generative adversarial networks (GANs) for image synthesis. By the end of this deep learning book, you’ll have learned the skills essential for building deep learning models with TensorFlow and Keras.
Table of Contents (9 chapters)
Preface

Bidirectional RNNs

The RNN models we've just looked at – LSTMs, GRUs – are powerful indeed and provide extremely good results when it comes to sequence-processing tasks. Now, let's discuss how to make them even more powerful, and the methods that yield the amazing successes in deep learning that you have been hearing about.

Let's begin with the idea of bidirectional RNNs. The idea applies to all variants of RNNs, including, but not limited to, LSTMs and GRUs. Bidirectional RNNs process the sequence in both directions, allowing the network to have both backward and forward information about the sequence, providing it with a much richer context:

Figure 6.17: Bidirectional LSTM

The bidirectional model essentially employs two RNNs in parallel – one as the "forward layer" and the other as the "backward layer". As shown in the preceding figure, the forward layer processes the sequence in the order of its...