verbrannter/handwriting_recognition_from_kaggle

该数据集包含来自慈善活动获取的超过四十万手写名字。数据集通过图像处理技术将扫描文档上的字符转换成数字格式，这种技术在处理机器打印字体时表现良好。然而，由于个人书写风格的多样性，机器在准确识别手写字符时仍然面临巨大挑战。数据集中共有第一名206,799个，姓氏207,024个。数据被划分为训练集（330,396个样本）、测试集（41,300个样本）和验证集（41,292个样本）。数据集已经清理了空或部分空的条目。

The dataset comprises over four hundred thousand handwritten names obtained from charitable initiatives. Character Recognition employs image processing techniques to transform characters present on scanned documents into digital formats. It generally exhibits good performance with machine-printed fonts. Nonetheless, machines still encounter formidable obstacles in accurately identifying handwritten characters due to the vast diversity in individual writing styles. The total number of first names was 206,799, while the total number of surnames was 207,024. The data was partitioned into a training set (330,396 samples), testing set (41,300 samples), and validation set (41,292 samples) respectively. The DATASET has been cleaned from empty or partly empty entries.