README.md

metadata

license: apache-2.0
size_categories:
  - 100K<n<1M
task_categories:
  - image-to-text
dataset_info:
  - config_name: default
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 392473380.05
        num_examples: 76318
    download_size: 383401054
    dataset_size: 392473380.05
  - config_name: full
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 392478490.025
        num_examples: 76319
      - name: validation
        num_bytes: 43364061.55
        num_examples: 8475
      - name: test
        num_bytes: 47643036.303
        num_examples: 9443
    download_size: 473618552
    dataset_size: 483485587.878
  - config_name: human_handwrite
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 16181778
        num_examples: 1200
      - name: validation
        num_bytes: 962283
        num_examples: 68
      - name: test
        num_bytes: 906906
        num_examples: 70
    download_size: 18056029
    dataset_size: 18050967
  - config_name: human_handwrite_print
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3152122.8
        num_examples: 1200
      - name: validation
        num_bytes: 182615
        num_examples: 68
      - name: test
        num_bytes: 181698
        num_examples: 70
    download_size: 1336052
    dataset_size: 3516435.8
  - config_name: small
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 261296
        num_examples: 50
      - name: validation
        num_bytes: 156489
        num_examples: 30
      - name: test
        num_bytes: 156489
        num_examples: 30
    download_size: 588907
    dataset_size: 574274
  - config_name: synthetic_handwrite
    features:
      - name: image
        dtype: image
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 496610333.066
        num_examples: 76266
      - name: validation
        num_bytes: 63147351.515
        num_examples: 9565
      - name: test
        num_bytes: 62893132.805
        num_examples: 9593
    download_size: 616418996
    dataset_size: 622650817.3859999
configs:
  - config_name: default
    data_files:
      - split: train
        path: full/train-*
  - config_name: full
    data_files:
      - split: train
        path: full/train-*
      - split: validation
        path: full/validation-*
      - split: test
        path: full/test-*
  - config_name: human_handwrite
    data_files:
      - split: train
        path: human_handwrite/train-*
      - split: validation
        path: human_handwrite/validation-*
      - split: test
        path: human_handwrite/test-*
  - config_name: human_handwrite_print
    data_files:
      - split: train
        path: human_handwrite_print/train-*
      - split: validation
        path: human_handwrite_print/validation-*
      - split: test
        path: human_handwrite_print/test-*
  - config_name: small
    data_files:
      - split: train
        path: small/train-*
      - split: validation
        path: small/validation-*
      - split: test
        path: small/test-*
  - config_name: synthetic_handwrite
    data_files:
      - split: train
        path: synthetic_handwrite/train-*
      - split: validation
        path: synthetic_handwrite/validation-*
      - split: test
        path: synthetic_handwrite/test-*
tags:
  - code

LaTeX OCR 的数据仓库

本数据仓库是专为 LaTeX_OCR 及 LaTeX_OCR_PRO 制作的数据，来源于 https://zenodo.org/record/56198#.V2p0KTXT6eA 以及 https://www.isical.ac.in/~crohme/ 以及我们自己构建。

如果这个数据仓库有帮助到你的话，请点亮 ❤️like ++

后续追加新的数据也会放在这个仓库 ~~

原始数据仓库在github LinXueyuanStdio/Data-for-LaTeX_OCR.

数据集

本仓库有 5 个数据集

1. small 是小数据集，样本数 110 条，用于测试
2. full 是印刷体约 100k 的完整数据集。实际上样本数略小于 100k，因为用 LaTeX 的抽象语法树剔除了很多不能渲染的 LaTeX。
3. synthetic_handwrite 是手写体 100k 的完整数据集，基于 full 的公式，使用手写字体合成而来，可以视为人类在纸上的手写体。样本数实际上略小于 100k，理由同上。
4. human_handwrite 是手写体较小数据集，更符合人类在电子屏上的手写体。主要来源于 CROHME。我们用 LaTeX 的抽象语法树校验过了。
5. human_handwrite_print 是来自 human_handwrite 的印刷体数据集，公式部分和 human_handwrite 相同，图片部分由公式用 LaTeX 渲染而来。

使用

加载训练集

• name 可选 small, full, synthetic_handwrite, human_handwrite, human_handwrite_print
• split 可选 train, validation, test

>>> from datasets import load_dataset
>>> train_dataset = load_dataset("linxy/LaTeX_OCR", name="small", split="train")
>>> train_dataset[2]["text"]
\rho _ { L } ( q ) = \sum _ { m = 1 } ^ { L } \ P _ { L } ( m ) \ { \frac { 1 } { q ^ { m - 1 } } } .
>>> train_dataset[2]
{'image': <PIL.PngImagePlugin.PngImageFile image mode=RGB size=200x50 at 0x15A5D6CE210>,
 'text': '\\rho _ { L } ( q ) = \\sum _ { m = 1 } ^ { L } \\ P _ { L } ( m ) \\ { \\frac { 1 } { q ^ { m - 1 } } } .'}
>>> len(train_dataset)
50

加载所有

>>> from datasets import load_dataset
>>> dataset = load_dataset("linxy/LaTeX_OCR", name="small")
>>> dataset
DatasetDict({
    train: Dataset({
        features: ['image', 'text'],
        num_rows: 50
    })
    validation: Dataset({
        features: ['image', 'text'],
        num_rows: 30
    })
    test: Dataset({
        features: ['image', 'text'],
        num_rows: 30
    })
})