---
title: "SciFigQual-Bench：基于全文语境的科学图像质量评估基准"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.27084v1"
permalink: "https://feed.astralor.com/items/itm_b159fea421e9a582be15de0d"
published: "2026-07-29T16:07:44.000Z"
score: 82
channel: "ai"
tags: ["GPT-5.6 Sol","SciFigQual-Bench","SFQ-Agent","基准数据集","科学图像质量评估","跨模态学习"]
---

# SciFigQual-Bench：基于全文语境的科学图像质量评估基准

## AI 点评

> 这篇论文精准击中了科研图像质量评估的空白点：现有方法要么只关注自然照片，要么仅做表面比较。SciFigQual-Bench不仅提供了大规模专家标注数据集，还绑定了完整论文上下文，使得评估更符合实际审稿场景。SFQ-Agent的分阶段跨模态设计也很有启发性——先从各个模态收集证据再融合，比直接端到端推理更可解释。对于学术出版平台、论文审稿系统以及研究科学可视化的团队来说，这是一个值得关注的基准。

## 速览

现有图像质量评估方法主要面向自然照片或AI生成内容，无法直接应用于科学论文中的图表。本文提出SciFigQual-Bench，一个包含6308张科学图像、覆盖2020-2025年顶级计算机会议论文的全文本基准，从清晰度、布局、标题匹配、上下文相关性及误导风险五个维度进行专家标注。同时设计SFQ-Agent分阶段跨模态评估框架，通过收集与融合模态证据实现可审计的精细评分。实验表明，基于GPT-5.6-Sol的SFQ-Agent (F3) 在eval1200子集上达到最低平均绝对误差(0.418)和最高一致率(93.4%)，显著优于直接评估和辅助VLM方案。

## 关键点

- 构建了包含6308张科学图像的全文本基准，由多位领域专家对五个维度独立评分，形成金标准标注。
- 每个图像绑定标题、引用句子和论文上下文，突破了以往仅视觉表面比较的局限。
- 提出SFQ-Agent分阶段跨模态评估框架，通过证据收集与融合实现可审计的精细评分。
- 基于GPT-5.6-Sol的SFQ-Agent (F3) 在评估中平均绝对误差仅0.418，一致率93.4%，优于直接评估与辅助VLM方法。

## 全文

Scientific images are the core elements of presenting experimental conclusions, elaborating system architecture, and supporting comparative arguments in scientific papers. However, existing image quality assessment (IQA) methods are predominantly designed for natural photographs or AI-generated content, which cannot be directly applied to scientific papers. The few existing studies on scholarly charts remain confined to visual-surface comparisons, failing to verify caption alignment, citation relevance, or visual misleadingness. To address this, we propose SciFigQual-Bench, a full-text contextual benchmark that evaluates scientific images across five dimensions (clarity, layout, caption fit, context relevance, and misleading risk). The data covers top computer-science conferences from 2020 to 2025; 6,308 images were independently scored by multiple domain experts in five dimensions and aggregated into gold-standard annotations. Unlike previous scientific figure benchmarks, our dataset binds each image to its caption, citing sentence, and manuscript context. To enable automated evaluation on this benchmark, we designed a staged cross-modal evaluation framework SFQ-Agent to achieve auditable and refined scoring through the collection and fusion of modal evidence. Multiple mainstream large models were evaluated on the test subset eval1200, and SFQ-Agent (F3) equipped with GPT-5.6-Sol achieved the lowest overall average absolute error (0.418) and the highest consistency rate (93.4%), consistently outperforming both direct evaluation and auxiliary (Sidecar) visual language model evaluation schemes.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_b159fea421e9a582be15de0d