Mobile Deep Learning with TensorFlow Lite, ML Kit and Flutter: Build scalable real-world projects to implement end-to-end neural networks on Android and iOS

Cover
Title Page
Copyright and Credits
About Packt
Contributors
Table of Contents
Preface
Chapter 01: Introduction to Deep Learning for Mobile
	Growth of AI-powered mobile devices
		Changes in hardware to support AI
			Why do mobile devices need to have AI chips?
			Improved user experience with AI on mobile devices
				Personalization
				Virtual assistants
				Facial recognition
				AI-powered cameras
				Predictive text
		Most popular mobile applications that use AI
			Netflix
			Seeing AI
			Allo
			English Language Speech Assistant
			Socratic
	Understanding machine learning and deep learning
		Understanding machine learning
		Understanding deep learning
			The input layer
			The hidden layers
			The output layer
			The activation function
	Introducing some common deep learning architectures
		Convolutional neural networks
		Generative adversarial networks
		Recurrent neural networks
		Long short-term memory
	Introducing reinforcement learning and NLP
		Reinforcement learning
		NLP
	Methods of integrating AI on Android and iOS
		Firebase ML Kit
		Core ML
		Caffe2
		TensorFlow
	Summary
Chapter 02: Mobile Vision - Face Detection Using On-Device Models
	Technical requirements
	Introduction to image processing
		Understanding images
		Manipulating images
			Rotation
			Grayscale conversion
	Developing a face detection application using Flutter
		Adding the pub dependencies
		Building the application
			Creating the first screen
				Building the row title
				Building the row with button widgets
				Creating the whole user interface
			Creating the second screen
				Getting the image file
				Analyzing the image to detect faces
				Marking the detected faces
				Displaying the final image on the screen
			Creating the final MaterialApp
	Summary
Chapter 03: Chatbot Using Actions on Google
	Technical requirements
	Understanding the tools available for creating chatbots
		Wit.ai
		Dialogflow
			How does Dialogflow work? 
	Creating a Dialogflow account
	Creating a Dialogflow agent
	Understanding the Dialogflow Console
		Creating an Intent and grabbing entities
	Creating your first action on Google
		Why would you want to build an action on Google?
	Creating Actions on a Google project
		Creating an integration to the Google Assistant
	Implementing a Webhook
	Deploying a webhook to Cloud Functions for Firebase
	Creating an Action on Google release
	Creating the UI for the conversational application
		Creating the Text Controller
		Creating ChatMessage
	Integrating the Dialogflow agent
	Adding audio interactions with the assistant
		Adding the plugin
		Adding SpeechRecognition
		Adding the mic button
	Summary
Chapter 04: Recognizing Plant Species
	Technical requirements
	Introducing image classification
	Understanding the project architecture
	Introducing the Cloud Vision API
	Configuring the Cloud Vision API for image recognition
		Enabling the Cloud Vision API
		Creating a Cloud Vision API key
	Using an SDK/tools to build a model
		Introducing Google's Colaboratory
	Creating a custom TensorFlow Lite model for image recognition
	Creating a Flutter application
		Choosing between two different models
		Creating the second screen
			Creating the user interface
			Adding the functionality
			Displaying the chosen image on the screen
	Running image recognition
		Using the Cloud Vision API
		Using an on-device TensorFlow Lite model
		Updating the UI with results
	Summary
Chapter 05: Generating Live Captions from a Camera Feed
	Designing the project architecture
	Understanding an image caption generator
		Understanding the dataset
		Building an image caption generation model
			Initializing the caption dataset
			Preparing the caption dataset
			Training
			Testing
		Creating a simple click-deploy image caption generation model
	Understanding the camera plugin
		Installing the camera plugin
		Adding methods for persistent storage and proper execution
		Coding
	Creating a camera application
		Building the camera preview
	Generating image captions from the camera feed
	Creating the material app
	Summary
Chapter 06: Building an Artificial Intelligence Authentication System
	Technical requirements
	A simple login application
		Creating the UI
	Adding Firebase authentication
		Creating auth.dart
		Adding authentication in SignupSigninScreen
		Creating the main screen
		Creating the home screen
		Creating main.dart
	Understanding anomaly detection for authentication
	A custom model for authenticating users
		Building a model for an authentication validity check
		Hosting the custom authentication validation model
	Implementing ReCaptcha for spam protection
		ReCAPTCHA v2
			Obtaining the API key
			Code integration
	Deploying the model in Flutter
	Summary
Chapter 07: Speech/Multimedia Processing - Generating Music Using AI
	Designing the project's architecture
	Understanding multimedia processing
		Image processing
		Audio processing
			Magenta
		Video processing 
	Developing RNN-based models for music generation
		Creating the LSTM-based model
		Deploying a model using Flask
	Deploying an audio generation API on Android and iOS
		Creating the UI
		Adding Audio Player
		Deploying the model
		Creating the final material app
	Summary
Chapter 08: Reinforced Neural Network-Based Chess Engine
	Introduction to reinforcement learning
	Reinforcement learning in mobile games
	Exploring Google's DeepMind
		AlphaGo
		Alpha Zero
		Monte Carlo tree search
	Alpha Zero-like AI for Connect 4
		Creating a virtual representation of the board
		Allowing moves according to the game's rules
		The state management system
		Facilitating gameplay
		Generating sample gameplays
		System training
		Monte Carlo tree search implementation
		Implementing the neural network
	Underlying project architecture
	Developing a GCP-hosted REST API for the chess engine
		Understanding the Universal Chess Interface
		Deployment on GCP
			Request for a quota increase on GPU instances
			Creating a GPU instance
			Deploying the script
	Creating a simple chess UI on Android
		Adding dependencies to pubspec.yaml
		Understanding the mapping structure
		Placing the images of the actual pieces
		Making the pieces movable
	Integrating the chess engine API with a UI
		Creating the material app
	Summary
Chapter 09: Building an Image Super-Resolution Application
	Basic project architecture
	Understanding GANs
	Understanding how image super-resolution works
		Understanding image resolution
			Pixel resolution
			Spatial resolution
			Temporal resolution
			Spectral resolution
			Radiometric resolution
		Understanding SRGANs
	Creating a TensorFlow model for super-resolution
		Project directory structure
		Creating an SRGAN model for super-resolution
	Building the UI for the application
	Getting pictures from the device's local storage
	Hosting a TensorFlow model on DigitalOcean
		Creating a Flask server script
		Deploying the Flask script to DigitalOcean Droplet
	Integrating a hosted custom model on Flutter
	Creating the Material app
	Summary
Chapter 10: Road Ahead
	Understanding recent trends in DL on mobile applications
		Math solver
		Netflix
		Google Maps
		Tinder
		Snapchat
	Exploring the latest developments in DL on mobile devices
		Google's MobileNet
		Alibaba Mobile Neural Network
	Exploring current research areas for DL in mobile apps
		Fashion images
		Self-Attention Generative Adversarial Networks
		Image animation
	Summary
Appendix
Other Books You May Enjoy
Index