DDIM
date
Feb 24, 2024
slug
ddim
status
Published
summary
Takeaways from DDIM
tags
DDIM
Diffusion model
Paper Reading
type
Post
I’m a visual learner, so it’s easier for me to learn a concept if they are visually structured in a meaningful way. Here’s my attempt with “Denoising Diffusion Implicit Models”, an influential paper that came out in ICLR 2021.
Takeaways
- DDIM actually has 2 contributions
- A variance hyper-parameter schedule is introduced to increase the flexibility of the inference process.
- when , it’s deterministic/implicit
- when , it’s equivalent to DDPM
- Realized that the variational objective doesn’t depend on the joint distribution , but rather on the marginals “independently”. This allows for a re-formulation of the inference and generation process to a non-markov process, while keeping the training process/objective unchanged (same model!)
- These 2 contributions are independent, in that the reformulation can be applied directly to DDPM without introducing the variance schedule.
Cheatsheet
