Self-Supervised Score-Based Despeckling for SAR Imagery via Log-Domain Transformation

2026-01-20

AI Summary

This work proposes a self-supervised despeckling method for synthetic aperture radar (SAR) images, which are degraded by multiplicative Gamma-distributed speckle noise that severely hinders subsequent analysis. The approach uniquely integrates a logarithmic transformation with a score-based generative model: the log-transform converts multiplicative speckle noise into approximately additive Gaussian noise, enabling the formulation of a self-supervised learning objective in the transformed domain using only a single noisy image—without requiring clean reference data. By leveraging this strategy, the method effectively suppresses speckle while preserving structural details and achieves significantly faster inference compared to existing self-supervised techniques, offering an efficient and robust solution for SAR image restoration.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

The speckle noise inherent in Synthetic Aperture Radar (SAR) imagery significantly degrades image quality and complicates subsequent analysis. Given that SAR speckle is multiplicative and Gamma-distributed, effectively despeckling SAR imagery remains challenging. This paper introduces a novel self-supervised framework for SAR image despeckling based on score-based generative models operating in the transformed log domain. We first transform the data into the log-domain and then convert the speckle noise residuals into an approximately additive Gaussian distribution. This step enables the application of score-based models, which are trained in the transformed domain using a self-supervised objective. This objective allows our model to learn the clean underlying signal by training on further corrupted versions of the input data itself. Consequently, our method exhibits significantly shorter inference times compared to many existing self-supervised techniques, offering a robust and practical solution for SAR image restoration.

Problem

• SAR despeckling
• speckle noise
• multiplicative noise
• image restoration
• Gamma-distributed noise

Innovation

• score-based generative models
• self-supervised learning
• log-domain transformation
• SAR despeckling
• additive Gaussian approximation