相关数据集
Salesforce/blip3-ocr-200m
BLIP3-OCR-200M数据集旨在解决当前视觉语言模型（VLMs）在处理和解释富含文本的图像（如文档和图表）时的局限性。该数据集在预训练阶段集成了光学字符识别（OCR）数据，通过提供详细的文本信息来增强视觉与语言的对齐。数据集专注于富含文本的图像，并包含OCR特定的注释，使得处理文档、图表等文本密集内容更加准确。数据集以Parquet格式存储，便于高效存储、处理和检索OCR元数据和图像。该数据
Hugging Face2025-02-03 更新680
OCR text derived from digitised books published 1820 - 1829. ALTO XML.
This set consists 2739 volumes, published between 1820-1829. The dataset comprises text from the collection of digitised books created using Optical Character Recognition (OCR) technology. The books
DataCite Commons2021-01-28 更新60
Data set of the paper "Publishing an OCR ground truth data set for reuse in an unclear copyright setting"
The data set consists of a METS file for each of the PDFs that were used for transcription and a directory data/page_xml that contains the transcriptions of the ground truth in PAGE-XML format. In par
NIAID Data Ecosystem50
AnnikaSimonsen/FO-OCRtrain
This is artificial Faroese OCR training data created by collection real OCR errors and inserting them into 38 million tokens of non-OCRed text. The parallel data is set up as a TSV file with the firs
Hugging Face2023-05-01 更新140
najwaerrachidy/layoutlmv3-cordv2_binary
该数据集包含图像、文字、框和标签等特征。数据集分为训练、验证和测试三个部分，其中训练集包含800个样本，验证集和测试集各包含100个样本。数据集的下载大小为2307636587字节，总大小为2314323029字节。
Hugging Face2024-07-11 更新60
