AI Disruption

AI Disruption

Share this post

AI Disruption
AI Disruption
The Key Noise Prediction Module in Diffusion Models Relies on UNet(AI Painting Creation Intro Course 8)
Copy link
Facebook
Email
Notes
More

The Key Noise Prediction Module in Diffusion Models Relies on UNet(AI Painting Creation Intro Course 8)

Explore UNet's role in image segmentation and AI art. Learn how it powers noise prediction in diffusion models and its unique applications across various fields.

Meng Li's avatar
Meng Li
Aug 22, 2024
∙ Paid
1

Share this post

AI Disruption
AI Disruption
The Key Noise Prediction Module in Diffusion Models Relies on UNet(AI Painting Creation Intro Course 8)
Copy link
Facebook
Email
Notes
More
1
Share

Welcome to the "AI Painting Creation Intro Course" Series

Table of Contents

Table of Contents

Meng Li
·
June 7, 2024
Read full story

Previously, we explored the diffusion model's noise-adding and denoising process and understood the basics of Transformers.

I also hinted earlier about using a UNet network to predict noise at each step.

Today, I'll explain the core concepts of UNet. We'll focus on answering these three questions:

  • How does the UNet model work?

  • What makes the UNet structure in various AI art models unique?

  • How can UNet be combined with Transformers?

By grasping these concepts, you'll be able to modify the noise prediction model in your work and lay a solid foundation for our hands-on training in diffusion models.

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2025 Meng Li
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share

Copy link
Facebook
Email
Notes
More