SiT: Exploring Flow and Diffusion-Based Generative Models with Scalable Interpolant Transformers

Nanye Ma, Mark Goldstein, Michael S. Albergo, Nicholas M. Boffi, Eric Vanden-Eijnden, Saining Xie

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present Scalable Interpolant Transformers (SiT), a family of generative models built on the backbone of Diffusion Transformers (DiT). The interpolant framework, which allows for connecting two distributions in a more flexible way than standard diffusion models, makes possible a modular study of various design choices impacting generative models built on dynamical transport: learning in discrete or continuous time, the objective function, the interpolant that connects the distributions, and deterministic or stochastic sampling. By carefully introducing the above ingredients, SiT surpasses DiT uniformly across model sizes on the conditional ImageNet 256×256 and 512×512 benchmark using the exact same model structure, number of parameters, and GFLOPs. By exploring various diffusion coefficients, which can be tuned separately from learning, SiT achieves an FID-50K score of 2.06 and 2.62, respectively. Code is available here: https://github.com/willisma/SiT.

Original languageEnglish (US)
Title of host publicationComputer Vision – ECCV 2024 - 18th European Conference, Proceedings
EditorsAleš Leonardis, Elisa Ricci, Stefan Roth, Olga Russakovsky, Torsten Sattler, Gül Varol
PublisherSpringer Science and Business Media Deutschland GmbH
Pages23-40
Number of pages18
ISBN (Print)9783031729799
DOIs
StatePublished - 2024
Event18th European Conference on Computer Vision, ECCV 2024 - Milan, Italy
Duration: Sep 29 2024Oct 4 2024

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume15135 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th European Conference on Computer Vision, ECCV 2024
Country/TerritoryItaly
CityMilan
Period9/29/2410/4/24

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'SiT: Exploring Flow and Diffusion-Based Generative Models with Scalable Interpolant Transformers'. Together they form a unique fingerprint.

Cite this