A comprehensive, self-contained guide to how machines learn to create images — from the first generative networks to the diffusion models and text-…

Welcome to Image Generation: From Pixels to Diffusion Models.

A comprehensive, self-contained guide to how machines learn to create images — from the first generative networks to the diffusion models and text-to-image systems behind today's AI art tools. Like its companion volume on machine learning, this book blends intuition, mathematics, and runnable code, building from first principles up to Stable Diffusion, ControlNet, and beyond.

This title is part of the ShriIra library and is free to read in full, right here — our small contribution to making world-class knowledge easy to reach.

A note on reading it: open the Contents menu at the top of the reader to jump between chapters, use the Aa menu to set a comfortable text size, theme (light, sepia, or night), and single- or two-page layout. Your place is saved automatically, so you can always pick up where you left off.

We hope it serves you well.

— Shriira Press

Contents

1. Chapter 1 — What Is Image Generation?
2. Chapter 2 — Images, Probability, and the Generative Problem
3. Chapter 3 — Neural Building Blocks for Generation
4. Chapter 4 — Autoencoders and Variational Autoencoders
5. Chapter 5 — Generative Adversarial Networks
6. Chapter 6 — Autoregressive and Discrete-Token Models
7. Chapter 7 — The Diffusion Idea: Forward and Reverse
8. Chapter 8 — Training and Sampling Diffusion Models
9. Chapter 9 — Conditioning and Guidance
10. Chapter 10 — Latent Diffusion and Modern Architectures
11. Chapter 11 — Anatomy of a Text-to-Image System
12. Chapter 12 — Control, Customization, and Editing
13. Chapter 13 — Evaluation, Ethics, and Deployment
14. Appendix A — Notation and Symbols
15. Appendix B — Further Reading