KLIP: localized distribution shift detection via KL-divergence with diffusion priors in Inverse Problems
This work addresses the critical problem of detecting subtle and localized distribution shifts in medical imaging for clinicians, which is an incremental improvement over existing OOD detection methods.
This paper introduces a new out-of-distribution (OOD) detection metric, KLIP, for inverse problems using diffusion priors. KLIP can detect both whole-image and localized distribution shifts without requiring knowledge of the shifted distribution or calibration data, successfully identifying subtle shifts like tumors in CT scans.
Diffusion models have shown promising performance as data-driven priors for computational imaging, as well as some capacity to detect out-of-distribution (OOD) images. However, existing approaches to OOD detection often require some knowledge of the shifted distribution, fail to detect subtle or localized distribution shifts, and operate on full images, rather than the indirect measurements available in inverse problems. We propose an OOD detection metric based on the Kullback-Leibler divergence between the diffusion prior and the posterior distribution, that (i) does not require any calibration data or knowledge of the shifted distribution, and (ii) can detect whole images as OOD as well as localize OOD patches within an image. Experimentally, we show that this metric can detect subtle yet semantically meaningful distribution shifts, such as the shift from healthy liver CT scans to those with tumors, and generalizes across different types of diffusion models, datasets, and inverse problems. Our code can be found at https://github.com/voilalab/KLIP.