Machine Learning

1 readers

1 users here now

Community Rules:

Be nice. No offensive behavior, insults or attacks: we encourage a diverse community in which members feel safe and have a voice.
Make your post clear and comprehensive: posts that lack insight or effort will be removed. (ex: questions which are easily googled)
Beginner or career related questions go elsewhere. This community is focused in discussion of research and new projects that advance the state-of-the-art.
Limit self-promotion. Comments and posts should be first and foremost about topics of interest to ML observers and practitioners. Limited self-promotion is tolerated, but the sub is not here as merely a source for free advertisement. Such posts will be removed at the discretion of the mods.

founded 1 year ago

MODERATORS

communick@academy.garden

[R] Beyond U: Making Diffusion Models Faster & Lighter (alien.top)

submitted 1 year ago by APaperADay@alien.top to c/machinelearning@academy.garden

5 comments fedilink hide all child comments

Paper: https://arxiv.org/abs/2310.20092

Abstract:

Diffusion models are a family of generative models that yield record-breaking performance in tasks such as image synthesis, video generation, and molecule design. Despite their capabilities, their efficiency, especially in the reverse denoising process, remains a challenge due to slow convergence rates and high computational costs. In this work, we introduce an approach that leverages continuous dynamical systems to design a novel denoising network for diffusion models that is more parameter-efficient, exhibits faster convergence, and demonstrates increased noise robustness. Experimenting with denoising probabilistic diffusion models, our framework operates with approximately a quarter of the parameters and 30% of the Floating Point Operations (FLOPs) compared to standard U-Nets in Denoising Diffusion Probabilistic Models (DDPMs). Furthermore, our model is up to 70% faster in inference than the baseline models when measured in equal conditions while converging to better quality solutions.

https://preview.redd.it/djk9mdlc9e0c1.png?width=995&format=png&auto=webp&s=65a002f1f320e68b71753ac32c6386c22e76c1c9

https://preview.redd.it/i87gizkc9e0c1.png?width=1108&format=png&auto=webp&s=34f25ecc319ffa34f545e850a5c95cb007e0abd8

you are viewing a single comment's thread
view the rest of the comments

[–] impossiblefork@alien.top 1 points 1 year ago (1 children)

No, it isn't. It's perfectly comprehensible.

The description of the architecture isn't incredibly clear, but it's enough to get the idea. I'd have liked to see the details, but if they want to write it like this that's fine.

[–] pm_me_your_pay_slips@alien.top 1 points 1 year ago (1 children)

Care to write a clear explanation of the method here?

[–] impossiblefork@alien.top 1 points 1 year ago

They use Neural ODEs as the denoiser, using multiple neural ODEs in a chsin and they somehow stick a time emedding into them.