Advanced Deep Learning with TensorFlow 2 and Keras - Second Edition

By : Rowel Atienza

Advanced Deep Learning with TensorFlow 2 and Keras - Second Edition

By: Rowel Atienza

Overview of this book

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.

Preface

Who this book is for

What this book covers

To get the most out of this book

Get in touch

Introducing Advanced Deep Learning with Keras

1. Why is Keras the perfect deep learning library?

2. MLP, CNN, and RNN

3. Multilayer Perceptron (MLP)

4. Convolutional Neural Network (CNN)

5. Recurrent Neural Network (RNN)

6. Conclusion

7. References

Free Chapter

Deep Neural Networks

1. Functional API

2. Deep Residual Network (ResNet)

3. ResNet v2

4. Densely Connected Convolutional Network (DenseNet)

5. Conclusion

6. References

Autoencoders

1. Principles of autoencoders

2. Building an autoencoder using Keras

3. Denoising autoencoders (DAEs)

4. Automatic colorization autoencoder

5. Conclusion

6. References

Generative Adversarial Networks (GANs)

1. An Overview of GANs

2. Implementing DCGAN in Keras

Improved GANs

2. Least-squares GAN (LSGAN)

3. Auxiliary Classifier GAN (ACGAN)

4. Conclusion

5. References

Disentangled Representation GANs

1. Disentangled representations

2. StackedGAN

4. Conclusion

5. References

Cross-Domain GANs

1. Principles of CycleGAN

2. Conclusion

3. References

Variational Autoencoders (VAEs)

1. Principles of VAE

2. Conditional VAE (CVAE)

3. 𝛽-VAE – VAE with disentangled latent representations

4. Conclusion

5. References

Deep Reinforcement Learning

1. Principles of Reinforcement Learning (RL)

2. The Q value

3. Q-learning example

4. Nondeterministic environment

5. Temporal-difference learning

6. Deep Q-Network (DQN)

7. Conclusion

8. References

Policy Gradient Methods

1. Policy gradient theorem

2. Monte Carlo policy gradient (REINFORCE) method

3. REINFORCE with baseline method

4. Actor-Critic method

5. Advantage Actor-Critic (A2C) method

6. Policy Gradient methods using Keras

7. Performance evaluation of policy gradient methods

Object Detection

3. Ground truth anchor boxes

4. Loss functions

5. SSD model architecture

6. SSD model architecture in Keras

7. SSD objects in Keras

8. SSD model in Keras

9. Data generator model in Keras

10. Example dataset

11. SSD model training

12. Non-Maximum Suppression (NMS) algorithm

13. SSD model validation

14. Conclusion

15. References

Semantic Segmentation

1. Segmentation

2. Semantic segmentation network

3. Semantic segmentation network in Keras

4. Example dataset

5. Semantic segmentation validation

6. Conclusion

7. References

Unsupervised Learning Using Mutual Information

1. Mutual Information

2. Mutual Information and Entropy

3. Unsupervised learning by maximizing the Mutual Information of discrete random variables

4. Encoder network for unsupervised clustering

5. Unsupervised clustering implementation in Keras

6. Validation using MNIST

7. Unsupervised learning by maximizing the Mutual Information of continuous random variables

8. Estimating the Mutual Information of a bivariate Gaussian

9. Unsupervised clustering using continuous random variables in Keras

10. Conclusion

11. References

Other Books You May Enjoy

Index

Customer Reviews

5 star

4 star

3 star

2 star

1 star

4. Loss functions

In SSD, there are thousands of anchor boxes. As discussed earlier in this chapter, the goal of object detection is to predict both the category and offsets of each anchor box. We can use the following loss functions for each prediction:

- Categorical cross-entropy loss for y_cls
- L1 or L2 for y_off. Note that only positive anchor boxes contribute to L1 is also known as mean absolute error (MAE) loss, while L2 is also known as mean squared error (MSE) loss.

The total loss function is:

For each anchor box, the network predicts the following:

y_cls or the category or class in the form of a one-hot vector
y_off = ((x_omin,y_omin),(x_omax,y_omax)) or the offsets in the form of pixel coordinates relative to anchor box.

For computational convenience, the offsets are better expressed in the form:

y_off = ((x_omin,y_omin),(x_omax,y_omax)) (Equation 11.4.2)

SSD is a supervised object detection algorithm...

Advanced Deep Learning with TensorFlow 2 and Keras - Second Edition

By : Rowel Atienza

Advanced Deep Learning with TensorFlow 2 and Keras - Second Edition

By: Rowel Atienza

Overview of this book

Related Content you might be interested in

Current Title:

Advanced Deep Learning with TensorFlow 2 and Keras - Second Edition

Hands-On Image Generation with TensorFlow

Hands-On Generative Adversarial Networks with Keras

Advanced Deep Learning with Python

4. Loss functions