A Sampling-Based Domain Generalization Study with Diffusion Generative Models
This addresses domain generalization for researchers in fields like astrophysics, offering a method to generate data in data-sparse scenarios, though it is incremental as it builds on existing diffusion models.
The paper tackled the problem of domain generalization for diffusion models by using a sampling-based approach with frozen pre-trained models, enabling synthesis of images from unseen domains without fine-tuning and maintaining original domain quality.
In this work, we investigate the domain generalization capabilities of diffusion models in the context of synthesizing images that are distinct from the training data. Instead of fine-tuning, we tackle this challenge from a sampling-based perspective using frozen, pre-trained diffusion models. Specifically, we demonstrate that arbitrary out-of-domain (OOD) images establish Gaussian priors in the latent spaces of a given model after inversion, and that these priors are separable from those of the original training domain. This OOD latent property allows us to synthesize new images of the target unseen domain by discovering qualified OOD latent encodings in the inverted noisy spaces, without altering the pre-trained models. Our cross-model and cross-domain experiments show that the proposed sampling-based method can expand the latent space and generate unseen images without impairing the generation quality of the original domain. We also showcase a practical application of our approach using astrophysical data, highlighting the potential of this generalization paradigm in data-sparse fields such as scientific exploration.