Diffusion Domain Expansion: Learning to Coordinate Pre-trained Diffusion Models
For practitioners using diffusion models, DDE enables scaling to larger domains without retraining, though the improvement is incremental over existing coordination methods.
Diffusion Domain Expansion (DDE) extends pre-trained diffusion models to generate larger objects and handle more complex conditioning, outperforming other coordinated generation methods in long audio track generation and conditional image generation.
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.