Packt+ | Advance your knowledge in tech

You're reading from Intelligent Mobile Projects with TensorFlow Build 10+ Artificial Intelligence apps using TensorFlow Mobile and Lite for iOS, Android, and Raspberry Pi

Product type Paperback

Published in May 2018

Publisher Packt

ISBN-13 9781788834544

Length 404 pages

Edition 1st Edition

Languages

Python

Tools

Android

Concepts

Artificial Intelligence

Author (1):

Tang

View More author details

Table of Contents (21) Chapters

Title Page

Dedication

Packt Upsell

Foreword

Contributors

Preface

1. Getting Started with Mobile TensorFlow

2. Classifying Images with Transfer Learning FREE CHAPTER

3. Detecting Objects and Their Locations

4. Transforming Pictures with Amazing Art Styles

5. Understanding Simple Speech Commands

6. Describing Images in Natural Language

7. Recognizing Drawing with CNN and LSTM

8. Predicting Stock Price with RNN

9. Generating and Enhancing Images with GAN

10. Building an AlphaZero-like Mobile Game App

11. Using TensorFlow Lite and Core ML on Mobile

12. Developing TensorFlow Apps on Raspberry Pi

1. Other Books You May Enjoy

Leave a review - let other readers know what you think

Index

Using the drawing classification model in Android

It's time to see how we can load and use the model in Android. In the previous chapters, we added TensorFlow support just by using the Android apps' build.gradle file and adding a line compile 'org.tensorflow:tensorflow-android:+' . Compared to iOS, where we have to build a custom TensorFlow library to fix different model loading or running errors (for example, inChapter 3, Detecting Objects and Their Locations, Chapter 4, Transforming Pictures with Amazing Art Styles, andChapter 5, Understanding Simple Speech Commands), the default TensorFlow library for Android has better support for registered operations and data types, which could be because Android is Google's first-class citizen while iOS is second, even a close second.

The fact is that, when we deal with all kinds of amazing models out there, it's just a matter of time before we have to face the inevitable: we have to build the TensorFlow library for Android manually to fix some errors...