TinyMind汉字书法识别数据集

TinyMind汉字书法识别的训练集和测试集，内有40000张汉字图片。训练集：每个汉字400张图片，共计40000张图片，训练集是标注好的数据，图片按照图片上的文字分类到不同的文件夹中，也就是生活文件夹的名字就是文件夹里面所有图片的标签。
测试集：
第一部分：每汉字100张图片共计10000张图片，供参赛人员测试算法模型
第二部分：每汉子50张以上图片共1643张图片，用来评测。
提交csv文件
文件第一列是测试集中图片文件的文件名，第二列是推断出来的图片上文字可能的五个汉字。

Training and test datasets for TinyMind Chinese character calligraphy recognition, containing a total of 40,000 Chinese character images.
Training set: 400 images per distinct Chinese character, summing up to 40,000 fully annotated images. Images are organized into separate folders categorized by the text content on them, where the folder name directly serves as the ground-truth label for all images stored within that folder.
Test set is split into two parts:
Part 1: 100 images per Chinese character, totaling 10,000 images, intended for participants to test their developed algorithm models.
Part 2: More than 50 images per Chinese character, with a total of 1,643 images, used for official model evaluation.
Submission requires a CSV file, in which the first column corresponds to the filename of the test set image, and the second column lists the top five predicted Chinese characters for the text on the corresponding image.

搜集汇总

背景与挑战

背景概述

TinyMind汉字书法识别数据集包含40000张训练图片和11643张测试图片，用于汉字书法识别任务。训练集按汉字分类，测试集分为两部分供算法测试和评测使用。