Skip to main content
  • Book
  • © 2021

Synthetic Data for Deep Learning

  • The first book about synthetic data, an important field which is rapidly rising in popularity throughout machine learning

  • Provides a wide survey of several different fields where synthetic data is or can potentially be useful, including domain adaptation and differential privacy

  • Contains a very extensive list of references, and in certain specific fields goes sufficiently in-depth to say that it discusses or at least mentions all relevant work

Part of the book series: Springer Optimization and Its Applications (SOIA, volume 174)

Buying options

eBook USD 89.00
Price excludes VAT (USA)
  • ISBN: 978-3-030-75178-4
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
Softcover Book USD 119.99
Price excludes VAT (USA)
Hardcover Book USD 169.99
Price excludes VAT (USA)

This is a preview of subscription content, access via your institution.

Table of contents (12 chapters)

  1. Front Matter

    Pages i-xii
  2. Introduction: The Data Problem

    • Sergey I. Nikolenko
    Pages 1-17
  3. Deep Learning and Optimization

    • Sergey I. Nikolenko
    Pages 19-58
  4. Deep Neural Networks for Computer Vision

    • Sergey I. Nikolenko
    Pages 59-95
  5. Generative Models in Deep Learning

    • Sergey I. Nikolenko
    Pages 97-137
  6. The Early Days of Synthetic Data

    • Sergey I. Nikolenko
    Pages 139-159
  7. Synthetic Data for Basic Computer Vision Problems

    • Sergey I. Nikolenko
    Pages 161-194
  8. Synthetic Simulated Environments

    • Sergey I. Nikolenko
    Pages 195-215
  9. Synthetic Data Outside Computer Vision

    • Sergey I. Nikolenko
    Pages 217-226
  10. Directions in Synthetic Data Development

    • Sergey I. Nikolenko
    Pages 227-234
  11. Synthetic-to-Real Domain Adaptation and Refinement

    • Sergey I. Nikolenko
    Pages 235-268
  12. Privacy Guarantees in Synthetic Data

    • Sergey I. Nikolenko
    Pages 269-283
  13. Promising Directions for Future Work

    • Sergey I. Nikolenko
    Pages 285-294
  14. Back Matter

    Pages 295-348

About this book

This is the first book on synthetic data for deep learning, and its breadth of coverage may render this book as the default reference on synthetic data for years to come. The book can also serve as an introduction to several other important subfields of machine learning that are seldom touched upon in other books. Machine learning as a discipline would not be possible without the inner workings of optimization at hand. The book includes the necessary sinews of optimization though the crux of the discussion centers on the increasingly popular tool for training deep learning models, namely synthetic data. It is expected that the field of synthetic data will undergo exponential growth in the near future. This book serves as a comprehensive survey of the field.  

In the simplest case, synthetic data refers to computer-generated graphics used to train computer vision models. There are many more facets of synthetic data to consider. In the section on basic computer vision, the book discusses fundamental computer vision problems, both low-level (e.g., optical flow estimation) and high-level (e.g., object detection and semantic segmentation), synthetic environments and datasets for outdoor and urban scenes (autonomous driving), indoor scenes (indoor navigation), aerial navigation, and simulation environments for robotics. Additionally, it touches upon applications of synthetic data outside computer vision (in neural programming, bioinformatics, NLP, and more). It also surveys the work on improving synthetic data development and alternative ways to produce it such as GANs.

The book introduces and reviews several different approaches to synthetic data in various domains of machine learning, most notably the following fields: domain adaptation for making synthetic data more realistic and/or adapting the models to be trained on synthetic data and differential privacy for generating synthetic data with privacy guarantees. This discussion is accompanied by an introduction into generative adversarial networks (GAN) and an introduction to differential privacy.

Keywords

  • synthetic data
  • deep learning
  • low-level computer vision
  • object detection
  • segmentation
  • GANs
  • domain transfer
  • machine learning models
  • deep learning and optimization
  • neural networks computer vision
  • synthetic simulated environment
  • computer vision problems

Authors and Affiliations

  • Synthesis AI, San Francisco, USA

    Sergey I. Nikolenko

About the author

Sergey I. Nikolenko is a computer scientist specializing in machine  learning and analysis of algorithms. He is the Head of AI at Synthesis  AI, a San Francisco based company specializing on the generation and use of synthetic data for modern machine learning models, and also serves as the Head of the Artificial Intelligence Lab at the Steklov Mathematical Institute at St. Petersburg, Russia. Dr. Nikolenko's interests include synthetic data in machine learning, deep learning models for natural language processing, image manipulation, and computer vision, and algorithms for networking. His previous research includes works on cryptography, theoretical computer science, and algebra.

Bibliographic Information

Buying options

eBook USD 89.00
Price excludes VAT (USA)
  • ISBN: 978-3-030-75178-4
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
Softcover Book USD 119.99
Price excludes VAT (USA)
Hardcover Book USD 169.99
Price excludes VAT (USA)