tqa
收藏资源简介:
这是一个从Textbook Question Answering (TQA)数据集派生出的数据集,专为ViDoRe基准测试而重新格式化了测试集的字段名。数据集中的`text_description`列包含了使用EasyOCR从图像中提取的OCR文本。此数据集是完整数据集的1000个随机行的子集。
This is a dataset derived from the Textbook Question Answering (TQA) dataset, with the field names of its test set reformatted specifically for the ViDoRe benchmark. The `text_description` column contains OCR text extracted from images using EasyOCR. This dataset is a subset of 1000 randomly selected rows from the full dataset.
TQA数据集概述
数据集基本信息
- 来源:基于Textbook Question Answering dataset (TQA)的测试集重构
- 用途:专为ViDoRe基准测试设计
- 样本量:1,000条随机抽样数据(完整数据集可访问https://huggingface.co/datasets/jinaai/tqa-test)
数据结构
特征字段
query:文本类型(string)image:图像类型(image)image_filename:文本类型(string)text_description:通过EasyOCR从图像提取的OCR文本(string)
数据划分
- 测试集(test):
- 样本数量:1,000例
- 数据大小:179.36MB
- 下载大小:160.92MB
学术引用
bibtex @inproceedings{Kembhavi2017TQA, title={Are You Smarter Than A Sixth Grader? Textbook Question Answering for Multimodal Machine Comprehension}, author={Kembhavi, Aniruddha and Salvatori, Tommaso and Kolve, Eric and Mottaghi, Roozbeh and Schwenk, Dustin and Farhadi, Ali and Yatskar, Mark}, booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, year={2017} }
版权声明
- 保留原始文档作者的所有权利
- 数据仅限研究及教育用途
- 侵权投诉联系:support-data (at) jina.ai




