PGDP5K
收藏资源简介:
PGDP5K是一个大规模的平面几何图形解析数据集,由中国科学院自动化研究所模式识别国家重点实验室创建。该数据集包含5000个图形样本,涵盖16种形状、5种位置关系、22种符号类型和6种文本类型,旨在通过精细的标注提高几何图形解析的准确性。创建过程中,研究团队通过分类、标注和关系构建等步骤,确保数据集的质量和多样性。该数据集主要应用于智能教育领域,特别是解决几何问题解析的挑战,推动相关算法的发展。
PGDP5K is a large-scale planar geometry parsing dataset developed by the State Key Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences. The dataset consists of 5000 graphic samples, covering 16 shape categories, 5 positional relationship types, 22 symbol types and 6 text types, aiming to improve the accuracy of geometry parsing via fine-grained annotations. During its construction, the research team ensured the quality and diversity of the dataset through procedures including classification, annotation and relationship construction. This dataset is mainly applied in the field of intelligent education, specifically addressing the challenges of geometry problem parsing and promoting the development of relevant algorithms.

- 1PGDP5K: A Diagram Parsing Dataset for Plane Geometry Problems中国科学院自动化研究所模式识别国家重点实验室 · 2022年



