How Diffusion Models Generate Images

Diffusion Models (DMs) generate images by denoising Gaussian noise through numerical simulations of Probability Flow ODEs or SDEs. The model is trained using denoising score matching, and classifier-free guidance enhances sampling quality for text-to-image generation.


This content originally appeared on HackerNoon and was authored by Synthesizing

:::info Authors:

(1) Dustin Podell, Stability AI, Applied Research;

(2) Zion English, Stability AI, Applied Research;

(3) Kyle Lacey, Stability AI, Applied Research;

(4) Andreas Blattmann, Stability AI, Applied Research;

(5) Tim Dockhorn, Stability AI, Applied Research;

(6) Jonas Müller, Stability AI, Applied Research;

(7) Joe Penna, Stability AI, Applied Research;

(8) Robin Rombach, Stability AI, Applied Research.

:::

Abstract and 1 Introduction

2 Improving Stable Diffusion

2.1 Architecture & Scale

2.2 Micro-Conditioning

2.3 Multi-Aspect Training

2.4 Improved Autoencoder and 2.5 Putting Everything Together

3 Future Work

\ Appendix

A Acknowledgements

B Limitations

C Diffusion Models

D Comparison to the State of the Art

E Comparison to Midjourney v5.1

F On FID Assessment of Generative Text-Image Foundation Models

G Additional Comparison between Single- and Two-Stage SDXL pipeline

References

C Diffusion Models

\ Sampling. In practice, this iterative denoising process explained above can be implemented through the numerical simulation of the Probability Flow ordinary differential equation (ODE) [47]

\

\ Classifier-free guidance. Classifier-free guidance [13] is a technique to guide the iterative sampling process of a DM towards a conditioning signal c by mixing the predictions of a conditional and an unconditional model

\ where w ≥ 0 is the guidance strength. In practice, the unconditional model can be trained jointly alongside the conditional model in a single network by randomly replacing the conditional signal c with a null embedding in Eq. (3), e.g., 10% of the time [13]. Classifier-free guidance is widely used to improve the sampling quality, trading for diversity, of text-to-image DMs [30, 38]

\

:::info This paper is available on arxiv under CC BY 4.0 DEED license.

:::

\


This content originally appeared on HackerNoon and was authored by Synthesizing


Print Share Comment Cite Upload Translate Updates
APA

Synthesizing | Sciencx (2024-10-03T19:04:29+00:00) How Diffusion Models Generate Images. Retrieved from https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/

MLA
" » How Diffusion Models Generate Images." Synthesizing | Sciencx - Thursday October 3, 2024, https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/
HARVARD
Synthesizing | Sciencx Thursday October 3, 2024 » How Diffusion Models Generate Images., viewed ,<https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/>
VANCOUVER
Synthesizing | Sciencx - » How Diffusion Models Generate Images. [Internet]. [Accessed ]. Available from: https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/
CHICAGO
" » How Diffusion Models Generate Images." Synthesizing | Sciencx - Accessed . https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/
IEEE
" » How Diffusion Models Generate Images." Synthesizing | Sciencx [Online]. Available: https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/. [Accessed: ]
rf:citation
» How Diffusion Models Generate Images | Synthesizing | Sciencx | https://www.scien.cx/2024/10/03/how-diffusion-models-generate-images/ |

Please log in to upload a file.




There are no updates yet.
Click the Upload button above to add an update.

You must be logged in to translate posts. Please log in or register.