Redefining Quality Criteria and Distance-Aware Score Modeling for Image Editing Assessment

TL;DR

This paper introduces DS-IEQA, a novel framework for image editing quality assessment that jointly learns evaluation criteria and score representations, addressing limitations of previous heuristic-based methods.

Contribution

It proposes Feedback-Driven Metric Prompt Optimization and Token-Decoupled Distance Regression Loss to improve alignment with human criteria and model score continuity.

Findings

Outperforms existing methods in image editing quality assessment

Ranks 4th in the 2026 NTIRE X-AIGC Quality Assessment Track 2

Effectively models score space structure and human criteria alignment

Abstract

Recent advances in image editing have heightened the need for reliable Image Editing Quality Assessment (IEQA). Unlike traditional methods, IEQA requires complex reasoning over multimodal inputs and multi-dimensional assessments. Existing MLLM-based approaches often rely on human heuristic prompting, leading to two key limitations: rigid metric prompting and distance-agnostic score modeling. These issues hinder alignment with implicit human criteria and fail to capture the continuous structure of score spaces. To address this, we propose Define-and-Score Image Editing Quality Assessment (DS-IEQA), a unified framework that jointly learns evaluation criteria and score representations. Specifically, we introduce Feedback-Driven Metric Prompt Optimization (FDMPO) to automatically refine metric definitions via probabilistic feedback. Furthermore, we propose Token-Decoupled Distance Regression Loss (TDRL), which decouples numerical tokens from language modeling to explicitly model score continuity through expected distance minimization. Extensive experiments show our method's superior performance; it ranks 4th in the 2026 NTIRE X-AIGC Quality Assessment Track 2 without any additional training data.