Practical Artificial Intelligence With Swift by Nugent, Tim -- Read -- Imperial Library of Trantor

Index

Preface

Resources Used in This Book Audience and Approach Organization of This Book Using This Book

Our Tasks

Conventions Used in This Book Using Code Examples O’Reilly Online Learning How to Contact Us Acknowledgments

I. Fundamentals and Tools 1. Artificial Intelligence!?

Practical AI with Swift…and Python?

Code Examples

Why Swift?

Why AI?

What Is AI and What Can It Do?

Deep Learning versus AI? Where Do the Neural Networks Come In? Ethical, Effective, and Appropriate Use of AI

Practical AI Tasks A Typical Task-Based Approach

2. Tools for Artificial Intelligence

Why Top Down? Great Tools for Great AI Tools from Apple

CoreML CreateML Turi Create Apple’s Other Frameworks CoreML Community Tools

Tools from Others

Swift for TensorFlow TensorFlow to CoreML Model Converter Other Converters

AI-Adjacent Tools

Python Keras, Pandas, Jupyter, Colaboratory, Docker, Oh My! Other People’s Tools

What’s Next?

3. Finding or Building a Dataset

Planning and Identifying Data to Target

Negation as Failure Closed-World Assumptions

Finding a Dataset

Where to Look What to Look Out for

Building a Dataset

Data Recording Data Collation Data Scraping

Preparing a Dataset

Getting to Know a Dataset Cleaning a Dataset Transforming a Dataset Verifying the Suitability of a Dataset

Apple’s Models

II. Tasks 4. Vision

Practical AI and Vision Task: Face Detection

Problem and Approach Building the App What Just Happened? How Does This Work? Improving the App Even More Improvements

Task: Barcode Detection Task: Saliency Detection Task: Image Similarity

Problem and Approach Building the App What Just Happened? How Does This Work? Next Steps

Task: Image Classification

Problem and Approach Building the App AI Toolkit and Dataset Incorporating the Model in the App Improving the App

Task: Drawing Recognition

Problem and Approach AI Toolkit and Dataset Building the App What’s Next?

Task: Style Classification

Converting the Model Using the Model

Next Steps

5. Audio

Audio and Practical AI Task: Speech Recognition

Problem and Approach Building the App What Just Happened? How Does This Work? What’s Next?

Task: Sound Classification

Problem and Approach Building the App AI Toolkit and Dataset Creating a Model Incorporating the Model in the App Improving the App

Next Steps

6. Text and Language

Practical AI, Text, and Language Task: Language Identification Task: Named Entity Recognition Task: Lemmatization, Tagging, and Tokenization

Parts of Speech Tokenizing a Sentence

Task: Sentiment Analysis

Problem and Approach Building the App AI Toolkit and Dataset Creating a Model Incorporating the Model in the App

Task: Custom Text Classifiers

AI Toolkit and Dataset

Next Steps

7. Motion and Gestures

Practical AI, Motion, and Gestures Task: Activity Recognition

Problem and Approach Building the App What Just Happened? How Does This Work?

Task: Gestural Classification for Drawing

Problem and Approach AI Toolkit and Dataset Building the App

Task: Activity Classification Problem and Approach

AI Toolkit and Dataset

Using the Model Task: Using Augmented Reality with AI Next Steps

8. Augmentation

Practical AI and Augmentation Task: Image Style Transfer

Problem and Approach Building the App AI Toolkit and Dataset Creating a Model Incorporating the Model in the App

Task: Sentence Generation

What Just Happened? How Does This Work?

Task: Image Generation with a GAN

Problem and Approach AI Toolkit and Dataset Building an App

Task: Recommending Movies

Problem and Approach AI Toolkit and Dataset Using a Recommender

Task: Regressor Prediction

Problem and Approach AI Toolkit and Dataset Using the Regressor in an App

Next Steps

9. Beyond Features

Task: Installing Swift for TensorFlow

Adding Swift for TensorFlow to Xcode Installing Swift for TensorFlow with Docker and Jupyter

Using Python with Swift Task: Training a Classifier Using Swift for TensorFlow Task: Using the CoreML Community Tools

The Problem The Process Using the Converted Model

On-Device Model Updates Task: Downloading Models On-device Next Steps

III. Beyond 10. AI and ML Methods

Terminology

AI/ML Components AI/ML Objectives Types of Values

Classification

Methods Applications

Clustering

Methods Applications

Next Steps

11. Looking Under the Hood

A Look Inside CoreML Vision

Face Detection Barcode Detection Saliency Detection Image Classification Image Similarity Bitmap Drawing Classification

Audio

Sound Classification Speech Recognition

Text and Language

Language Identification Named Entity Recognition Lemmatization, Tagging, Tokenization

Recommendations Prediction Text Generation Generation The Future of CoreML Next Steps

12. The Hard Way

Behind CoreML’s Magic Task: Building XOR

The Shape of Our Network

The Code

Building It Up Making It Work Tearing It Down Using the Neural Network Approximations of XOR

Training Next Steps

Index

← Prev
Back
Next →

← Prev
Back
Next →