English

Diffusion Domain Expansion: Learning to Coordinate Pre-trained Diffusion Models

Machine Learning 2026-05-25 v1

Abstract

In this paper, we propose Diffusion Domain Expansion (DDE), a method that efficiently extends pre-trained diffusion models to generate larger objects and handle more complex conditioning beyond their original capabilities. Our method employs a compact trainable network designed to coordinate the denoised outputs of pre-trained diffusion models. We demonstrate that the coordinator can be universally simple while being capable of generalizing to domains larger than those observed during its training time. We evaluate DDE on long audio track generation and conditional image generation, demonstrating its applicability across domains. DDE outperforms other approaches to coordinated generation with diffusion models in qualitative and quantitative evaluations.

Keywords

Cite

@article{arxiv.2605.23275,
  title  = {Diffusion Domain Expansion: Learning to Coordinate Pre-trained Diffusion Models},
  author = {Egor Lifar and Semyon Savkin and Timur Garipov and Shangyuan Tong and Tommi Jaakkola},
  journal= {arXiv preprint arXiv:2605.23275},
  year   = {2026}
}

Comments

Accepted as poster at ICML 2024 Workshop on Structured Probabilistic Inference and Generative Modeling (SPIGM)