# Diffusion-Prior Split Gibbs Sampling for Synthetic Aperture Radar Imaging under Incomplete Measurements

arXiv: https://arxiv.org/abs/2512.02768 · HTML: https://mlsanity.com/paper/2512.02768/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Inverse problems and image restoration, one of 14 directions in Diffusion models.
- Builds on 13 of its 67 references that are in the map.
- First posted on arXiv 2 Dec 2025.

## Placed in

- Diffusion models / [Inverse problems and image restoration](https://mlsanity.com/diffusion/inverse-problems-and-image-restoration/)

## Builds on (most cited first, 13 of 13)

- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [Generative Modeling by Estimating Gradients of the Data Distribution](https://mlsanity.com/paper/1907.05600/) (2019)
- [Diffusion Posterior Sampling for General Noisy Inverse Problems](https://mlsanity.com/paper/2209.14687/) (2022)
- [Fast Sampling of Diffusion Models with Exponential Integrator](https://mlsanity.com/paper/2204.13902/) (2022)
- [Understanding Diffusion Models: A Unified Perspective](https://mlsanity.com/paper/2208.11970/) (2022)
- [Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction](https://mlsanity.com/paper/2403.17042/) (2024)
- [Split-and-augmented Gibbs sampler - Application to large-scale inference problems](https://mlsanity.com/paper/1804.05809/) (2018)
- [Plug-and-Play split Gibbs sampler: embedding deep generative priors in Bayesian inference](https://mlsanity.com/paper/2304.11134/) (2023)
- [DiC: Rethinking Conv3x3 Designs in Diffusion Models](https://mlsanity.com/paper/2501.00603/) (2024)
- [ATASI-Net: An Efficient Sparse Reconstruction Network for Tomographic SAR Imaging with Adaptive Threshold](https://mlsanity.com/paper/2211.16855/) (2022)

## Built on by (most cited first, 1 of 1)

- [Learning to Replace MCMC in Split-Gibbs Diffusion Posterior Sampling via Deep Unfolding](https://mlsanity.com/paper/2609.30539/) (2026)

## Abstract (arXiv)

Synthetic aperture radar (SAR) imaging plays a critical role in all-weather, day-and-night remote sensing, yet reconstruction is often challenged by noise, undersampling, and complex scattering scenarios. Conventional methods, including matched filtering and sparsity-based compressed sensing, are limited in capturing intricate scene structures and frequently suffer from artifacts, elevated sidelobes, and loss of fine details. Recent diffusion models have demonstrated superior capability in representing high-order priors; however, existing diffusion-based SAR methods still yield degraded reconstructions due to oversimplified likelihood approximations in guided sampling. In this work, we propose a diffusion-driven split Gibbs sampling framework for SAR reconstruction, rigorously integrating measurement fidelity with learned diffusion priors. By alternately performing likelihood- and prior-driven updates via proximal sampling, this method ensures progressive convergence toward the true posterior while fully leveraging the expressive power of diffusion priors. Extensive experiments on simulated and Sentinel-1A datasets demonstrate substantial performance improvements: over 7 dB average PSNR gain in simulations, along with significant sidelobe suppression (MPLSR +2.96 dB, MISLR +11.5 dB) with respect to the best baseline result. On real-world Sentinel-1A data, the method achieves an average PSNR gain of 1.6 dB while effectively reducing artifacts and preserving scene details, including ridges, edges, and fine textures. These results underscore the potential of the adapted framework as a robust and generalizable solution for high-fidelity SAR imaging across diverse sensing scenarios.