howard-hou/OCR-VQA
收藏资源简介:
--- dataset_info: features: - name: image dtype: image - name: image_id dtype: string - name: questions sequence: string - name: answers sequence: string - name: ocr_tokens sequence: string - name: ocr_info list: - name: word dtype: string - name: bounding_box struct: - name: width dtype: float64 - name: height dtype: float64 - name: top_left_x dtype: float64 - name: top_left_y dtype: float64 - name: title dtype: string - name: authorName dtype: string - name: genre dtype: string - name: image_width dtype: int64 - name: image_height dtype: int64 - name: image_url dtype: string - name: set_name dtype: string splits: - name: train num_bytes: 7503971854.0 num_examples: 166022 - name: test num_bytes: 928616409.0 num_examples: 20796 - name: validation num_bytes: 920236957.0 num_examples: 20731 download_size: 2329997099 dataset_size: 9352825220.0 --- # Dataset Card for "OCR-VQA" [More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
数据集概述
数据集特征
- image: 图像数据
- image_id: 字符串类型
- questions: 字符串序列
- answers: 字符串序列
- ocr_tokens: 字符串序列
- ocr_info: 列表类型,包含:
- word: 字符串类型
- bounding_box: 结构体类型,包含:
- width: 浮点数类型
- height: 浮点数类型
- top_left_x: 浮点数类型
- top_left_y: 浮点数类型
- title: 字符串类型
- authorName: 字符串类型
- genre: 字符串类型
- image_width: 整数类型
- image_height: 整数类型
- image_url: 字符串类型
- set_name: 字符串类型
数据集分割
- train: 166022个样本,占用7503971854字节
- test: 20796个样本,占用928616409字节
- validation: 20731个样本,占用920236957字节
数据集大小
- 下载大小: 2329997099字节
- 数据集总大小: 9352825220字节



