arXiv 2511.114102025-11-14cs.CV

Q-Doc：多模态大语言模型中文档图像质量评估能力的基准测试

Q-Doc: Benchmarking Document Image Quality Assessment Capabilities in Multi-modal Large Language Models

South China Normal University（华南师范大学）；Nanyang Technological University（南洋理工大学）；City University of Hong Kong（香港城市大学）；Shenzhen Academy of Inspection and Quarantine（深圳检验检疫局）；Shenzhen University（深圳大学）

Jiaxi Huang, Dongxu Wu, Hanwei Zhu, Lingyu Zhu, Jun Xing, Xu Wang, Baoliang Chen

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对多模态大语言模型（MLLMs）文档图像质量评估（DIQA）能力未被充分探索的问题，提出三层评估框架Q-Doc，发现MLLMs存在相关局限性，且思维链（CoT）提示可显著提升其性能，为该领域提供基准与改进方向。

摘要：

多模态大语言模型（MLLMs）的快速发展使其能力拓展到高级视觉任务之外。然而，它们在文档图像质量评估（DIQA）方面的潜力仍未得到充分探索。为填补这一空白，我们提出Q-Doc，一个三层评估框架，用于系统地从粗、中、细三个粒度级别探测MLLMs的DIQA能力：a）在粗粒度级别，我们指导MLLMs为文档图像分配质量分数，并分析其与质量标注的相关性；b）在中粒度级别，我们设计失真类型识别任务，包括针对多失真场景的单选和多选测试；c）在细粒度级别，我们引入失真严重程度评估，让MLLMs对照人工标注参考对失真强度进行分类。我们的评估表明，虽然MLLMs具备初步的DIQA能力，但存在关键局限性：评分不一致、失真识别错误和严重程度判断失误。重要的是，我们发现思维链（CoT）提示能显著提升所有级别上的性能。我们的工作为MLLMs的DIQA能力提供了基准，揭示了其质量感知方面的明显缺陷以及有前景的改进路径。该基准和代码可在https://github.com/cydxf/Q-Doc公开获取。

英文摘要：

The rapid advancement of Multi-modal Large Language Models (MLLMs) has expanded their capabilities beyond high-level vision tasks. Nevertheless, their potential for Document Image Quality Assessment (DIQA) remains underexplored. To bridge this gap, we propose Q-Doc, a three-tiered evaluation framework for systematically probing DIQA capabilities of MLLMs at coarse, middle, and fine granularity levels. a) At the coarse level, we instruct MLLMs to assign quality scores to document images and analyze their correlation with Quality Annotations. b) At the middle level, we design distortion-type identification tasks, including single-choice and multi-choice tests for multi-distortion scenarios. c) At the fine level, we introduce distortion-severity assessment where MLLMs classify distortion intensity against human-annotated references. Our evaluation demonstrates that while MLLMs possess nascent DIQA abilities, they exhibit critical limitations: inconsistent scoring, distortion misidentification, and severity misjudgment. Significantly, we show that Chain-of-Thought (CoT) prompting substantially enhances performance across all levels. Our work provides a benchmark for DIQA capabilities in MLLMs, revealing pronounced deficiencies in their quality perception and promising pathways for enhancement. The benchmark and code are publicly available at: https://github.com/cydxf/Q-Doc.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。