QualiRAG: Retrieval-Augmented Generation for Visual Quality Understanding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cao, Linhan, Sun, Wei, Zhang, Weixia, Zhu, Xiangyang, Zhang, Kaiwei, Jia, Jun, Zhu, Dandan, Zhai, Guangtao, Min, Xiongkuo
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911399087702016
author Cao, Linhan
Sun, Wei
Zhang, Weixia
Zhu, Xiangyang
Zhang, Kaiwei
Jia, Jun
Zhu, Dandan
Zhai, Guangtao
Min, Xiongkuo
author_facet Cao, Linhan
Sun, Wei
Zhang, Weixia
Zhu, Xiangyang
Zhang, Kaiwei
Jia, Jun
Zhu, Dandan
Zhai, Guangtao
Min, Xiongkuo
contents Visual quality assessment (VQA) is increasingly shifting from scalar score prediction toward interpretable quality understanding -- a paradigm that demands \textit{fine-grained spatiotemporal perception} and \textit{auxiliary contextual information}. Current approaches rely on supervised fine-tuning or reinforcement learning on curated instruction datasets, which involve labor-intensive annotation and are prone to dataset-specific biases. To address these challenges, we propose \textbf{QualiRAG}, a \textit{training-free} \textbf{R}etrieval-\textbf{A}ugmented \textbf{G}eneration \textbf{(RAG)} framework that systematically leverages the latent perceptual knowledge of large multimodal models (LMMs) for visual quality perception. Unlike conventional RAG that retrieves from static corpora, QualiRAG dynamically generates auxiliary knowledge by decomposing questions into structured requests and constructing four complementary knowledge sources: \textit{visual metadata}, \textit{subject localization}, \textit{global quality summaries}, and \textit{local quality descriptions}, followed by relevance-aware retrieval for evidence-grounded reasoning. Extensive experiments show that QualiRAG achieves substantial improvements over open-source general-purpose LMMs and VQA-finetuned LMMs on visual quality understanding tasks, and delivers competitive performance on visual quality comparison tasks, demonstrating robust quality assessment capabilities without any task-specific training. The code will be publicly available at https://github.com/clh124/QualiRAG.
format Preprint
id arxiv_https___arxiv_org_abs_2601_18195
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle QualiRAG: Retrieval-Augmented Generation for Visual Quality Understanding
Cao, Linhan
Sun, Wei
Zhang, Weixia
Zhu, Xiangyang
Zhang, Kaiwei
Jia, Jun
Zhu, Dandan
Zhai, Guangtao
Min, Xiongkuo
Computer Vision and Pattern Recognition
Visual quality assessment (VQA) is increasingly shifting from scalar score prediction toward interpretable quality understanding -- a paradigm that demands \textit{fine-grained spatiotemporal perception} and \textit{auxiliary contextual information}. Current approaches rely on supervised fine-tuning or reinforcement learning on curated instruction datasets, which involve labor-intensive annotation and are prone to dataset-specific biases. To address these challenges, we propose \textbf{QualiRAG}, a \textit{training-free} \textbf{R}etrieval-\textbf{A}ugmented \textbf{G}eneration \textbf{(RAG)} framework that systematically leverages the latent perceptual knowledge of large multimodal models (LMMs) for visual quality perception. Unlike conventional RAG that retrieves from static corpora, QualiRAG dynamically generates auxiliary knowledge by decomposing questions into structured requests and constructing four complementary knowledge sources: \textit{visual metadata}, \textit{subject localization}, \textit{global quality summaries}, and \textit{local quality descriptions}, followed by relevance-aware retrieval for evidence-grounded reasoning. Extensive experiments show that QualiRAG achieves substantial improvements over open-source general-purpose LMMs and VQA-finetuned LMMs on visual quality understanding tasks, and delivers competitive performance on visual quality comparison tasks, demonstrating robust quality assessment capabilities without any task-specific training. The code will be publicly available at https://github.com/clh124/QualiRAG.
title QualiRAG: Retrieval-Augmented Generation for Visual Quality Understanding
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.18195