HierText
收藏arXiv2025-09-30 收录
官方服务：
资源简介：
该数据集主要关注手写样本的公开OCR训练数据，规模各异，其中大约有1.3千个手写单词的样本。该数据集的任务是用于OCR训练。
This dataset focuses on public OCR training data for handwritten samples, with varying scales, and contains approximately 1.3 thousand handwritten word samples. Its core task is designed for OCR training.
