When No-Reference Image Quality Models Meet MAP Estimation in Diffusion Latents

TL;DR

This paper introduces a novel approach that integrates no-reference image quality assessment models into a MAP estimation framework for image enhancement in diffusion latent space, enabling perceptual optimization and comparison of different models.

Contribution

It is the first to use NR-IQA models as priors in diffusion latent MAP estimation, improving real-world image enhancement and providing a new way to compare NR-IQA models.

Findings

Enhanced real-world images with complex distortions using the proposed method.

Different NR-IQA models produce varied enhanced outputs, enabling comparative analysis.

The combined approach yields better image quality while maintaining fidelity.

Abstract

Contemporary no-reference image quality assessment (NR-IQA) models can effectively quantify perceived image quality, often achieving strong correlations with human perceptual scores on standard IQA benchmarks. Yet, limited efforts have been devoted to treating NR-IQA models as natural image priors for real-world image enhancement, and consequently comparing them from a perceptual optimization standpoint. In this work, we show -- for the first time -- that NR-IQA models can be plugged into the maximum a posteriori (MAP) estimation framework for image enhancement. This is achieved by performing gradient ascent in the diffusion latent space rather than in the raw pixel domain, leveraging a pretrained differentiable and bijective diffusion process. Likely, different NR-IQA models lead to different enhanced outputs, which in turn provides a new computational means of comparing them. Unlike conventional correlation-based measures, our comparison method offers complementary insights into the respective strengths and weaknesses of the competing NR-IQA models in perceptual optimization scenarios. Additionally, we aim to improve the best-performing NR-IQA model in diffusion latent MAP estimation by incorporating the advantages of other top-performing methods. The resulting model delivers noticeably better results in enhancing real-world images afflicted by unknown and complex distortions, all preserving a high degree of image fidelity.