arXiv cs.AI· Radha Krishna Deshpande, Ajoy Mondal, Akash Manna, C V Jawahar·· 14 小时前AI 评分31
面向立法档案的鲁棒跨语言 RAG：如何跨越 OCR 噪声读取历史文献
Reading Beyond OCR Noise: Robust Cross-Lingual RAG for Legislative Archives
AI 导读
研究团队提出一套端到端框架，将基于视觉的版面分割、多脚本 OCR 与智能体式 RAG 流水线结合，用迭代智能体循环绕开 OCR 噪声和版面歧义后再生成答案。该系统在 20 万页的 Punjab Vidhan Sabha 档案上完成评测，对历史文档退化表现稳健，优于纯文本基线。
正文 · 原文
Abstract:India's historical public archives present severe computer vision and document image analysis challenges: decades of parliamentary records survive only as degraded, multi-column scanned pages featuring mixed scripts, complex layouts, and dense visual noise where OCR failure is the primary obstacle to information retrieval. We present an end-to-end framework that couples vision-based layout segmentation and multi-script OCR with an agentic retrieval-augmented generation (RAG) pipeline. Our approach specifically addresses how visual degradation and layout fragmentation propagate into text extraction errors, using an iterative agentic loop to navigate around OCR noise and layout ambiguities before generating answers. Evaluated on the 200,000-page Punjab Vidhan Sabha archive, our system demonstrates robust handling of historical document degradation, outperforming standard text-only baselines.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR) |
| Cite as: | arXiv:2610.35240 [cs.CV] |
| (or arXiv:2610.35240v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35240 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: RadhaKrishna Deshpande [view email]
[v1]
Mon, 26 Oct 2026 11:24:16 UTC (6,934 KB)
来源：arXiv cs.AI · arxiv.org