Restore, Assess, Repeat: A Unified Framework for Iterative Image Restoration

TL;DR

The paper introduces RAR, a unified iterative framework combining image quality assessment and restoration in the latent domain, achieving state-of-the-art results in handling various image degradations efficiently.

Contribution

It presents a novel end-to-end trainable model that integrates IQA and IR for adaptive, efficient image restoration across diverse degradation types.

Findings

Consistent improvements on single, unknown, and composite degradations.

State-of-the-art performance in image restoration tasks.

Efficient in latency and information preservation.

Abstract

Image restoration aims to recover high quality images from inputs degraded by various factors, such as adverse weather, blur, or low light. While recent studies have shown remarkable progress across individual or unified restoration tasks, they still suffer from limited generalization and inefficiency when handling unknown or composite degradations. To address these limitations, we propose RAR, a Restore, Assess and Repeat process, that integrates Image Quality Assessment (IQA) and Image Restoration (IR) into a unified framework to iteratively and efficiently achieve high quality image restoration. Specifically, we introduce a restoration process that operates entirely in the latent domain to jointly perform degradation identification, image restoration, and quality verification. The resulting model is fully trainable end to end and allows for an all-in-one assess and restore approach that dynamically adapts the restoration process. Also, the tight integration of IQA and IR into a unified model minimizes the latency and information loss that typically arises from keeping the two modules disjoint, (e.g. during image and/or text decoding). Extensive experiments show that our approach consistent improvements under single, unknown and composite degradations, thereby establishing a new state-of-the-art.