Twin-T 与 TwintVQA:一种可靠的结构-细节分离 VLM 及面向图表与表格任务的综合基准
Twin-T & TwintVQA: A Reliable Structure-Detail Separating VLM and a Comprehensive Benchmark for Chart and Table Tasks
双头图像编码器分离结构线索与细节的思路可迁移到其他需要同时理解布局和细粒度内容的视觉任务(如文档分析、场景文本理解);MINT 偏好学习方法强调数值与关键词保真度,为提升 VLM 在数值敏感任务上的可靠性提供了可复用的训练策略。
原文依据 we propose a dual-head image encoder that can separate structural cues and fine details from input images
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论随着视觉语言模型(VLM)的快速发展,对结构化视觉数据的自动分析需求日益增长。图表和表格通过规整的布局、明确的数字以及图表特有的阅读模式来承载定量信息,然而当前 VLM 仍未充分利用这些特性,常常导致数值错误和不可靠的分析。为克服这些局限,我们提出 Twin-T,一个两阶段专家 VLM,用于跨图像、LaTeX 和 Python 的全面图表-表格任务。在第一阶段,我们提出一种双头图像编码器,能够从输入图像中分离结构线索和细节信息。在第二阶段,我们提出 MINT,一种偏好学习方法,强调数值和关键词保真度以及视觉-文本匹配。此外,我们引入一个全面的 TwintVQA 基准,包含 17 种图表类型、11 种任务类型、3 种数据格式以及短/中/长问答设置。我们的模型缩小了开源与闭源模型在主流图表-表格基准上的差距,在超越开源模型的同时,甚至与 GPT-4o 和 Gemini-2.5-Pro 保持竞争力。
英文原文摘要 1126 字符
With the rapid development of Vision-Language Models (VLMs), there is a growing demand for automatic analysis of structured visual data. Charts and tables carry quantitative information through regular layouts, explicit numbers, and chart-specific reading patterns, yet current VLMs still underuse these properties, often causing value errors and unreliable analysis. To overcome these limitations, we propose Twin-T, a two-stage expert VLM for comprehensive chart-table tasks across Image, LaTeX, and Python. In stage 1, we propose a dual-head image encoder that can separate structural cues and fine details from input images. In stage 2, we propose MINT, a preference learning method that emphasizes numerical and keyword fidelity, as well as vision-text matching. Furthermore, we introduce a comprehensive TwintVQA benchmark with 17 chart types, 11 task types, 3 data formats, and short / medium / long QA settings. Our model narrows the gap between open-source and closed-source models on mainstream chart-table benchmarks, outperforming open-source models while even remaining competitive with GPT-4o and Gemini-2.5-Pro.