Asked about Generating by removing noise
How do diffusion models work?
An answer that works
A fixed forward process gradually adds noise to data. A model learns to reverse one step by predicting the noise added. Generation starts from pure noise and applies the learned reverse step repeatedly.
Reading an answer is not being able to give one. The section that teaches this is Generating by removing noise, 4 hours of Diffusion & generative models.
Learn this properlyAlso asked about Diffusion & generative models
Why did diffusion displace GANs?What is latent diffusion?How do you trade quality against latency in a diffusion model?What is classifier-free guidance?How would you add a specific style to a diffusion model?Write down the diffusion training objective and explain it.You trained a diffusion model and samples look like blurry blobs. What might be wrong?You need image generation under one second per image. What are your options?