Panoptic Pairwise Distortion Graph

2026-04-13

AI Summary

Existing methods for image-pair quality assessment struggle to model the structured relationships of region-level distortions, lacking fine-grained detail and interpretability. This work introduces a novel task—“distortion graphs”—by extending scene graphs to image pairs, where graph structures encode distortion types, severity levels, comparative relations, and quality scores across regions. To support this, we present PandaSet, the first region-level distortion dataset, along with PandaBench, a dedicated benchmark, and propose Panda, a specialized neural architecture that integrates graph modeling, regional visual analysis, and multimodal learning. Experiments demonstrate that Panda significantly outperforms existing approaches on PandaBench. While mainstream multimodal large language models initially exhibit limited performance, their capacity for region-level distortion understanding markedly improves when trained on or prompted with distortion graphs.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

In this work, we introduce a new perspective on comparative image assessment by representing an image pair as a structured composition of its regions. In contrast, existing methods focus on whole image analysis, while implicitly relying on region-level understanding. We extend the intra-image notion of a scene graph to inter-image, and propose a novel task of Distortion Graph (DG). DG treats paired images as a structured topology grounded in regions, and represents dense degradation information such as distortion type, severity, comparison and quality score in a compact interpretable graph structure. To realize the task of learning a distortion graph, we contribute (i) a region-level dataset, PandaSet, (ii) a benchmark suite, PandaBench, with varying region-level difficulty, and (iii) an efficient architecture, Panda, to generate distortion graphs. We demonstrate that PandaBench poses a significant challenge for state-of-the-art multimodal large language models (MLLMs) as they fail to understand region-level degradations even when fed with explicit region cues. We show that training on PandaSet or prompting with DG elicits region-wise distortion understanding, opening a new direction for fine-grained, structured pairwise image assessment.

Problem

• image quality assessment
• region-level degradation
• distortion understanding
• pairwise image comparison
• structured representation

Innovation

• Distortion Graph
• region-level assessment
• structured image comparison
• PandaSet
• multimodal large language models