IIIT-AR-13K
收藏资源简介:
IIIT-AR-13K数据集是由国际信息技术研究所视觉信息中心创建,专注于商业文档中的图形对象检测,特别是年度报告。该数据集包含13000个手动标注的页面,涵盖五种不同类别的图形对象:表格、图表、自然图像、标志和签名。数据集的创建涉及从多个公司和多种语言的公开年度报告中随机选择文档,并手动标注图形对象的边界框。IIIT-AR-13K数据集的应用领域包括商业文档和科技文章中的图形对象检测,旨在提高自动处理商业文档的效率和准确性。
The IIIT-AR-13K dataset was developed by the Visual Information Center at the International Institute of Information Technology, focusing on graphical object detection in business documents, particularly annual reports. This dataset contains 13,000 manually annotated pages, covering five categories of graphical objects: tables, charts, natural images, logos, and signatures. The construction of the dataset involves randomly selecting documents from public annual reports of multiple companies across various languages, followed by manual annotation of bounding boxes for graphical objects. Application scenarios of the IIIT-AR-13K dataset include graphical object detection in business documents and scientific articles, aiming to improve the efficiency and accuracy of automated business document processing.

- 1IIIT-AR-13K: A New Dataset for Graphical Object Detection in Documents国际信息技术研究所视觉信息中心 · 2020年



