A Single-Parameter Factor-Graph Image Prior

2026-01-13

AI Summary

This work addresses the lack of adaptive yet concise prior models in image denoising and contrast enhancement by proposing a factor-graph-based piecewise-smooth image prior. The model employs a Normal with Unknown Precision (NUP) prior, which adaptively captures local image structures using only a single global parameter, thereby achieving both expressive power and simplicity. Efficient optimization is realized through Gaussian message passing combined with the conjugate gradient method. Extensive experiments demonstrate that the proposed approach achieves state-of-the-art performance in both denoising and contrast enhancement tasks, confirming its effectiveness and practical utility.

Technology Category

• Reasoning under Uncertainty: Graphical Models
• Computer Vision: Learning & Optimization for CV
• Machine Learning: Probabilistic Circuits and Graphical Models

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We propose a novel piecewise smooth image model with piecewise constant local parameters that are automatically adapted to each image. Technically, the model is formulated in terms of factor graphs with NUP (normal with unknown parameters) priors, and the pertinent computations amount to iterations of conjugate-gradient steps and Gaussian message passing. The proposed model and algorithms are demonstrated with applications to denoising and contrast enhancement.

Problem

• image prior
• piecewise smooth
• factor graph
• image denoising
• contrast enhancement

Innovation

• factor graph
• NUP prior
• piecewise smooth image model
• Gaussian message passing
• adaptive image prior