ZaNioxX/DocILE_10_5_ImageClassification_donut

该数据集用于图像分类任务，主要包含图像和对应的标签。标签包括多种文档类型，如信用票据、借记票据、订单、形式发票、采购订单、收据、销售订单、税务发票和公用事业账单。数据集分为测试集和训练集，测试集包含21483个样本，训练集包含85939个样本。数据集的总大小为20064495900.858字节，下载大小为12741489204字节。

This dataset is intended for image classification tasks, primarily comprising images and their corresponding labels. The labels encompass various document types, including credit notes, debit notes, orders, proforma invoices, purchase orders, receipts, sales orders, tax invoices, and utility bills. The dataset is split into a test set and a training set, with 21483 samples in the test set and 85939 samples in the training set. The total size of the dataset is 20064495900.858 bytes, and the download size is 12741489204 bytes.

原始信息汇总

数据集概述

数据集配置

• 配置名称: default
• 数据文件:
  • 测试集: data/test-*
  • 训练集: data/train-*

数据集信息

• 特征:

  • image: 图像数据
  • label: 分类标签
    • 类别名称:
      • 0: credit_note
      • 1: debit_note
      • 2: order
      • 3: proforma
      • 4: purchase_order
      • 5: receipt
      • 6: sales_order
      • 7: tax_invoice
      • 8: utility_bill
  • ground_truth: 字符串类型

• 分割:

  • 测试集:
    • 字节数: 4160197623.858
    • 样本数: 21483
  • 训练集:
    • 字节数: 15904298277.0
    • 样本数: 85939

• 下载大小: 12741489204

• 数据集大小: 20064495900.858