让AI像人一样有选择地看长文档图像，更准更快还少瞎编。
InSight-doc: Agentic Visual Perception for Long-Document Understanding
- 用动态分辨率策略，只在关键区域高精度分析，节省计算资源。
- 在长文档上准确率提升4.3%~16.4%，幻觉减少超40%，推理快41%~68%。
- 适合需要处理复杂图文长文档的场景，如法律、医疗、金融领域。
长文档理解常需对大量视觉信息丰富的页面进行推理，导致计算成本高且易出现上下文遗忘。本文提出InSight-doc，一种将视觉分辨率作为可变推理资源的智能体视觉感知框架。该框架从低分辨率开始，仅对关键区域进行高分辨率聚焦以获取细粒度证据，无需依赖外部检索器。为训练该智能体，我们构建了包含17.9K高质量SFT样本和19.2K困难强化学习样本的主动感知语料库，其中包含区域级缩放轨迹。通过SFT+RL联合训练，InSight-doc-8B在文档VQA基准上相较基线提升4.3–16.4个百分点准确率；在长文档上，幻觉减少超过40%，推理延迟降低41%–68%，同时保持准确率优势。代码、数据集及模型已开源。
原文摘要 · Abstract (English)
Long-document understanding often requires reasoning over many visually rich pages, making inference costly and prone to context rot. In this work, we propose InSight-doc, an agentic visual perception framework that treats visual resolution as an adaptive reasoning-time resource. InSight-doc starts from low resolution and selectively zooms into high-resolution regions for finer evidence, without relying on any external retriever. To train such an agent, we construct an active-perception corpus of 17.9K high-quality SFT examples with region-level zoom-in trajectories, accompanied by 19.2K hard RL examples. Through SFT+RL, InSight-doc-8B improves the baseline by 4.3--16.4 accuracy points over document VQA benchmarks. On long documents, it reduces hallucination by more than 40% and inference latency by 41%--68% while maintaining an accuracy lead. Our code, datasets, and model are released at https://github.com/m-Just/InSight-doc .
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。