Skip to content
Preprint

Atlases Are Already Inside: Recovering Population Templates from Pretrained Diffusion Models

Sep 2026 · 0 citations · 47 references
Computer Science

TL;DR

A new inference-time sampler for diffusion models that gives a pretrained model a capability it was never trained for: constructing the atlas of the population it synthesizes, which is the intrinsic atlas of the population it synthesizes.

Abstract

We present a new inference-time sampler for diffusion models that gives a pretrained model a capability it was never trained for: constructing the atlas of the population it synthesizes. The sampler converges from every random seed to the population's central anatomy, which we call the \emph{intrinsic atlas}. The advantage is threefold. (1) It requires no retraining. A diffusion model that has already learned a coherent population, including the released ones, yields its atlas in a single inference pass without involving deformable registration. (2) It applies to multiple domains, such as brain MRI, chest X-ray, faces, and 3D shapes. (3) It extends to subpopulations. One age-conditioned model gives an atlas at any age in its training range, and the resulting family reproduces the CSF expansion of healthy aging. Evaluated as a registration target, the intrinsic atlas is best or second-best on every dataset against classical and learned templates, and the most central template on held-out brain MRI cohorts. Atlas construction can be reframed as a byproduct of generative modeling: a diffusion model is a learned representation of population structure, and the atlas is what it already contains.

View source

Similar papers

Sep 2026

UnfoldCRF: Structured Mask Refinement with Image-Conditioned Latent Regions

UnfoldCRF treats refinement as inference in a conditional random field over pixel labels and latent region variables, and zeroing individual messages shows where the corrections come from: the pairwise messages mostly fix boundaries, the region messages mostly fix non-boundary errors.

Chun-Ming He, Rihan Zhang, Lei Xu et al. · 1 citation
Preprint Sep 2026

Diffuse2Seg: Diffusion Models Can Segment Anything Without Supervision

Diffuse2Seg is introduced, which repurposes generative diffusion models for automatic mask generation by propagating a grid of point prompts through their self-attention representations in an edge-preserving manner, and shows that a model trained on Diffuse2Seg labels provides a strong initialization for semi-supervise...

Christoph Hümmer, Joachim Sicking, Fabian Hüger et al. · 0 citations
#artificial intelligence Preprint Oct 2026

Correction Training for Diffusion Models

The method substantially improves upon previously reported FID scores for unguided generation and, notably, unconditional generation, where producing high-quality samples without guidance has been particularly challenging.

Donghoon Ahn, Run-Qian Wang, Alexei A. Efros et al. · 0 citations
#machine learning Preprint Oct 2026

Neural Fields Encode Adaptation Geometry

Neural fields are usually evaluated by how well they reconstruct an observation. We show that this misses two useful properties of a fitted network: how easily it can adapt to new observations, and what its weights retain from earlier ones. We study these properties as adaptation geometry. For images, we meta-learn cla...

Prateik Sinha, Stefania Druga · 0 citations
Preprint Oct 2026

DART: Scaling a Shape Representation Across Anatomical Classes and Imaging Modalities

Current state-of-the-art segmentation models still produce anatomically implausible shapes, such as broken vessels and fragmented glands, on the structures that are hardest to segment. Learned shape representations capture what structures look like and can help correct these errors, but most cover a fixed set of anatom...

Koushik Swarna, P. Bassi, L. Ngo et al. · 0 citations
Preprint Oct 2026

D-MOD: Diffusion mode estimation for accelerated MRI

D-MOD, a fast algorithm that instead provides a posterior mode for MRI reconstruction, uses a minorize-maximize algorithm on a lower bound of the smoothed log-posterior, which provably increases the objective.

Md Kawsar Ahmed, Mathews Jacob · 1 citation

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.