Exploring Deepfakes: Deploy powerful AI techniques for face replacement and more with this comprehensive guide

by Bryan Lyon, Matt Tora

Length: 192 pages
Edition: 1
Language: English
Publisher: Packt Publishing
Publication Date: 2023-03-28
ISBN-10: 1801810699
ISBN-13: 9781801810692
Sales Rank: #2554955 (See Top 100 Books)

0 ratings

Print Book Look Inside

Master the innovative world of deepfakes and generative AI for face replacement with this full-color guide

Purchase of the print or Kindle book includes a free PDF eBook

Key Features

Understand what deepfakes are, their history, and how to use the technology ethically
Get well-versed with the workflow and processes involved to create your own deepfakes
Learn how to apply the lessons and techniques of deepfakes to your own problems

Book Description

Applying Deepfakes will allow you to tackle a wide range of scenarios creatively.

Learning from experienced authors will help you to intuitively understand what is going on inside the model. You’ll learn what deepfakes are and what makes them different from other machine learning techniques, and understand the entire process from beginning to end, from finding faces to preparing them, training the model, and performing the final swap.

We’ll discuss various uses for face replacement before we begin building our own pipeline. Spending some extra time thinking about how you collect your input data can make a huge difference to the quality of the final video. We look at the importance of this data and guide you with simple concepts to understand what your data needs to really be successful.

No discussion of deepfakes can avoid discussing the controversial, unethical uses for which the technology initially became known. We’ll go over some potential issues, and talk about the value that deepfakes can bring to a variety of educational and artistic use cases, from video game avatars to filmmaking.

By the end of the book, you’ll understand what deepfakes are, how they work at a fundamental level, and how to apply those techniques to your own needs.

What you will learn

Gain a clear understanding of deepfakes and their creation
Understand the risks of deepfakes and how to mitigate them
Collect efficient data to create successful deepfakes
Get familiar with the deepfakes workflow and its steps
Explore the application of deepfakes methods to your own generative needs
Improve results by augmenting data and avoiding overtraining
Examine the future of deepfakes and other generative AIs
Use generative AIs to increase video content resolution

Who this book is for

This book is for AI developers, data scientists, and anyone looking to learn more about deepfakes or techniques and technologies from Deepfakes to help them generate new image data. Working knowledge of Python programming language and basic familiarity with OpenCV, Pillow, Pytorch, or Tensorflow is recommended to get the most out of the book.

Cover
Preface
Title Page
Copyright
Contributors
Table of Contents
Part 1: Understanding Deepfakes
Chapter 1: Surveying Deepfakes
	Introducing deepfakes
	Exploring the uses of deepfakes
		Entertainment
		Parody
		Education
		Advertisements
	Discovering how deepfakes work
		Generative auto-encoders
	Assessing the limitations of generative AI
		Resolution
		Training required for each face pair
		Training data
	Looking at existing deepfake software
		Faceswap
		DeepFaceLab
		First Order Model
		Reface
	Summary
Chapter 2: Examining Deepfake Ethics and Dangers
	The unethical origin of deepfakes
	Being an ethical deepfaker
		Consent
		Respect
		Deception
		Putting it into practice
	The dangers of deepfakes
		Reputation
		Politics
		Avoiding consequences by claiming manipulation
	Preventing damage from deepfakes
		Starving the model of data
		Authenticating any genuine media
		Deepfake detection
		Public relations
		Public awareness
	Summary
Chapter 3: Acquiring and Processing Data
	Why data is important
	Understanding the value of variety
		Pose
		Expression
		Lighting
		Bringing this variety together
	Sourcing data
		Filming your own data
		Getting data from historical sources
	Improving your data
		Linear color
		Data matching
	Upscaling
	Summary
Chapter 4: The Deepfake Workflow
	Technical requirements
	Identifying suitable candidates for a swap
	Preparing the training images
		Extracting faces from your source data
		Curating training images
	Training a model
		Setting up
		Launching and monitoring training
		Manual intervention
	Applying a trained model to perform a swap
		The alignments file
		Cleaning the alignments file
		Fixing the alignments file
		Using the Preview tool
		Generating the swap
	Summary
Part 2: Getting Hands-On with the Deepfake Process
Chapter 5: Extracting Faces
	Technical requirements
	Getting image files from a video
	Running extract on frame images
		face_alignments.json
		face_bbox_{filename}_{face number}.png
		face_aligned_{filename}_{face number}.png
		face_mask_{filename}_{face number}.png
	Getting hands-on with the code
		Initialization
		Image preparation
		Face detection
		Face landmarking/aligning
	Summary
	Exercises
Chapter 6: Training a Deepfake Model
	Technical requirements
	Understanding convolutional layers
	Getting hands-on with AI
		Defining our upscaler
		Creating the encoder
		Building the decoders
	Exploring the training code
		Creating our models
		Looping over the training
		Teaching the network
		Saving results
	Summary
	Exercises
Chapter 7: Swapping the Face Back into the Video
	Technical requirements
	Preparing to convert video
	Getting hands-on with the convert code
		Initialization
		Loading the AI
		Preparing data
		The conversion loop
	Creating the video from images
	Summary
	Exercises
Part 3: Where to Now?
Chapter 8: Applying the Lessons of Deepfakes
	Technical requirements
	Aligning other types of images
		Finding an aligner
		Using the library
		Using the landmarks to align
	The power of masking images
		Types of masking
		Finding a usable mask for your object
		Examining an example
	Getting data under control
		Defining your rules
		Evolving your rules
		Dealing with errors
	Summary
Chapter 9: The Future of Generative AI
	Generating text
		Recent developments
		Building sentences
		The future of text generation
	Improving image quality
		Various tactics
		The future of image quality upgrading
	Text-guided image generation
		CLIP
		Image generation with CLIP
		The future of image generation
	Generating sound
		Voice swapping
		Text-guided music generation
		The future of sound generation
	Deepfakes
		Sound generation
		Text-guided image generation
		Improving image quality
		Text generation
		The future of deepfakes
		The future of AI ethics
	Summary
Index
About Packt
Other Books You May Enjoy

AI & Machine Learning Artificial Intelligence Data Processing Intelligence & Semantics Neural Networks

Donate to keep this site alive

To access the Link, solve the captcha.

How to download source code?

1. Go to: https://github.com/PacktPublishing

2. In the Find a repository… box, search the book title: Exploring Deepfakes: Deploy powerful AI techniques for face replacement and more with this comprehensive guide, sometime you may not get the results, please search the main title.