Steerable Conditional Diffusion for Out-of-Distribution Adaptation in Imaging Inverse Problems

• DOI: 10.48550/arxiv.2308.14409
• Publication Date: 2023-01-01
• Authors: Riccardo Barbano; Alexander Denker; Hyungjin Chung; Tae Hoon Roh; Simon Arrdige; Peter Maaß; Bangti Jin; Jong Chul Ye
• Keywords: Leverage (statistics); Hallucinating

Denoising diffusion models have emerged as the go-to framework for solving inverse problems in imaging. A critical concern regarding these is their performance on out-of-distribution (OOD) tasks, which remains an under-explored challenge. Realistic reconstructions inconsistent with measured data can be generated, hallucinating image features that are uniquely present training dataset. To simultaneously enforce data-consistency and leverage data-driven priors, we introduce a novel sampling called Steerable Conditional Diffusion. This adapts denoising network specifically to available data. Utilising our proposed method, achieve substantial enhancements OOD across diverse imaging modalities, advancing robust deployment of real-world applications.