Tag
该论文提出实体级分析作为补充视角来评估历史文档OCR质量，使用Berrutti档案馆175份文档及5,209条人工标注实体，比较了序列识别OCR系统、OCR专用视觉-语言模型和通用视觉-语言模型，发现不同评价标准下OCR排名会变化。
A project lasting a year and a half collected Bhutanese historical literature, published the four-volume 'Bhutan Ancient Sites Records', and open-sourced the related database.
AI companies are suspected of buying and destroying rare books to use as training data, raising concerns among booksellers and book lovers. The article contrasts this with the Internet Archive's careful, human-driven scanning process and Google's non-destructive patent as alternatives.