How far have we gone in Generative Image Restoration? A study on its capability, limitations and evaluation practices

TL;DR

This study systematically evaluates modern generative image restoration models across multiple quality dimensions, revealing performance disparities, evolving failure modes, and introducing a new perceptual quality assessment model.

Contribution

It provides the first large-scale, multi-dimensional evaluation pipeline for GIR models and uncovers a paradigm shift in their failure modes and capabilities.

Findings

Diffusion-based models outperform GANs in detail and semantic correctness.

A new IQA model aligns better with human perceptual judgments.

Identifies a shift from under-generation to over-generation issues.

Abstract

Generative Image Restoration (GIR) has achieved impressive perceptual realism, but how far have its practical capabilities truly advanced compared with previous methods? To answer this, we present a large-scale study grounded in a new multi-dimensional evaluation pipeline that assesses models on detail, sharpness, semantic correctness, and overall quality. Our analysis covers diverse architectures, including diffusion-based, GAN-based, PSNR-oriented, and general-purpose generation models, revealing critical performance disparities. Furthermore, our analysis uncovers a key evolution in failure modes that signifies a paradigm shift for the perception-oriented low-level vision field. The central challenge is evolving from the previous problem of detail scarcity (under-generation) to the new frontier of detail quality and semantic control (preventing over-generation). We also leverage our benchmark to train a new IQA model that better aligns with human perceptual judgments. Ultimately, this work provides a systematic study of modern generative image restoration models, offering crucial insights that redefine our understanding of their true state and chart a course for future development.