No. 420 of 500 No. 68 in Hands-On AI & Programming

Practical Deep Learning for Cloud, Mobile, and Edge

Real-World AI and Computer-Vision Projects Using Python, Keras, and TensorFlow

By Anirudh Koul, Siddha Ganju and Meher Kasam

Print length
620 pages
Publisher
O'Reilly Media
Reading level
Intermediate
ISBN-13
978-1492034865
Published
November 26, 2019
Written for
Software engineers and hobbyists who want to ship computer vision features on phones, browsers or edge devices.

Synopsis

Anirudh Koul, Siddha Ganju and Meher Kasam wrote this O'Reilly guide for developers who want to move from a deep learning idea to a working product. Rather than starting with theory, the book builds image classifiers with Keras and TensorFlow, then uses transfer learning to retrain them on small custom datasets in minutes. From there it widens into similarity search with embeddings, data pipelines and performance tuning, serving models from the cloud, and running them inside browsers, iOS apps with Core ML, Android apps with TensorFlow Lite, and edge boards such as the Raspberry Pi, Jetson Nano and Google Coral. Later chapters simulate an autonomous car in a game environment and train a miniature vehicle with reinforcement learning.

The approach is project-first: each chapter pairs runnable code from the authors' GitHub repository with practical checklists on accuracy, speed and debugging, plus dozens of short industry case studies. Readers come away knowing how to choose a model that fits a device's memory and latency budget, how to scale inference, and where real-world deployments fail. The tooling has moved on since 2019, and the book predates the transformer and generative AI wave, so its lasting value lies in the engineering habits rather than the exact library calls.

Scorecard

Editorial score

How it scored

Editorial score: 64.3 / 1005 weighted criteria

  • Influence30%

    18.0/30
  • Authority25%

    17.5/25
  • Practical value20%

    14.0/20
  • Currency15%

    6.8/15
  • Readability10%

    8.0/10
  • Editorial score

    64.3/100

How the rank was computed

Rank No. 420 of 500 comes from a composite of 49.1: observable signals 20.2 of 55 points plus editorial review 28.9 of 45 points.

All 11 inputs
InputPoints
Observable signals20.2/55
Reach (Amazon ratings volume)8.0/16.5
Consensus across best-of lists5.5/13.8
Awards and bestseller evidence0.0/8.3
Reader rating, sample-adjusted6.4/8.3
Amazon sales rank0.3/5.5
Recency0.0/2.8
Editorial review28.9/45
Author authority7.9/11.3
Influence8.1/13.5
Practical value6.3/9.0
Readability3.6/4.5
Currency3.0/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

A hands-on computer vision guide whose deployment lessons hold up, even though its 2019-era Keras and TensorFlow code has aged.

Strengths

  • Runnable code repository maintained alongside the book
  • Covers cloud, browser, mobile and edge deployment
  • Practical checklists for accuracy and speed tuning

Keep in mind

Code targets 2019-era Keras and TensorFlow releases and predates transformers and generative AI, so readers must adapt examples to current libraries.

Key takeaways

  1. Transfer learning lets a developer retrain a pretrained image classifier on a small custom dataset in minutes instead of training from scratch.
  2. Choosing a model for phones or edge boards means trading accuracy against memory, latency and power, and the book shows how to measure each.
  3. Many accuracy and speed problems trace back to data pipelines and training configuration, so systematic checklists often beat swapping in a bigger architecture.
  4. Embeddings from a trained network can power reverse image search and similarity features without labeling every image in a collection.

Notable facts

  • The authors' official code repository states that the book is featured as a learning resource on the official Keras website. Source (opens in a new tab)
  • NVIDIA's author page for co-author Siddha Ganju states that the roughly 600-page book is being translated into five languages. Source (opens in a new tab)

About the author

Anirudh Koul, Siddha Ganju and Meher Kasam

Anirudh Koul founded Microsoft's Seeing AI app for blind and low-vision users, later worked at Pinterest and served as machine learning lead at NASA's Frontier Development Lab; Carnegie Mellon gave him a 2020 Alumni Achievement Award. Co-author Siddha Ganju, a Forbes 30 Under 30 honoree, has worked at NVIDIA on self-driving systems and large language models. Meher Kasam is an iOS engineer who has worked at Square and Microsoft.

Questions

Is Practical Deep Learning for Cloud, Mobile, and Edge worth reading in 2026?
Yes for its deployment thinking, with caveats. The book's lessons on choosing models for limited hardware, tuning pipelines and serving inference remain sound, but its code reflects 2019-era Keras and TensorFlow releases and it does not cover transformers, large language models or generative AI. Readers should expect to update library calls and pair it with a newer resource on foundation models.
Who should read Practical Deep Learning for Cloud, Mobile, and Edge?
The book suits software engineers, mobile developers, data scientists and makers who already write Python and want to put computer vision models into real products. It assumes little deep learning theory, so it works as a first applied text, but readers who want mathematical depth on neural networks will need a separate textbook.
What projects does Practical Deep Learning for Cloud, Mobile, and Edge include?
Projects include an image classifier retrained with transfer learning, a reverse image search engine built on embeddings, a Not Hotdog style iOS app using Core ML, an Android app with TensorFlow Lite, browser demos, object detection, edge builds on Raspberry Pi, Jetson Nano and Google Coral, a self-driving simulation in a game engine, and a miniature racing car trained with reinforcement learning.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.