Image · Text / AI & Machine Learning
Document OCR&Parsing Data | 10 Million Pages | Computer Vision Data
Advertised by providerLicensing rights not independently verifiedPaid licence
Overview
Document OCR&Parsing Data | 10 Million Pages | Computer Vision Data is an image and text dataset. Typical applications include LLM pre-training and document understanding & OCR. Relevant sectors: AI & Machine Learning and Education. The listing mentions transcription. It is offered as a paid dataset; pricing details are provided on request.
Primary applications
What the listing offers
- Listing mentions transcription.
Million Level Document OCR and Structured Analysis Data, Including textbooks, e-books, teaching aids, patents, theses, etc. The annotated files include OCR annotations and structured analysis. Data Size 10 Million Pages Data Types Textbooks, Chinese E-books, Teaching Reference Books, Paper, Chinese Journals, English Journals Data Format The original document file format is PDF, the document image file format is. png, the OCR annotation file format is JSON, and the structured parsing file format is markdown(Tables and formulas are in Latex format or screenshot links) Data Accuracy If the transcription of the text is basically accurate, the markings are basically aligned, and there are no obvious typos, it is considered correct annotation. Divided by punctuation marks, the number of correctly annotated sentences should not be less than 90%.
Technical specifications
- Modalities
- Image, Text
Related datasets
More in Text & Language →Image · TextAdvertised
Natural Scenes & Document Photograph Scenes & Electronic Scenes OCR Data of 21 Countries | 500,000 Images | Computer Vision Data
500,000 Images - Natural Scenes & Document Photograph Scenes & Electronic Scenes OCR Data of 21 Countries.
- Language
- —
- Volume
- 500k images
- Domain
- AI & Machine Learning +1
- Use case
- LLM pre-training
USD 10,000
More text & language datasets
All 776 text & language datasets →- Documentary Video Dataset | 101,413 Hours | Large Language Model(LLM) Data
- Domain-Specific Benchmark Dataset
- Dutch Average Tone Speech Synthesis Corpus | 2 People | Text-to-Speech(TTS) Data
- Dutch Scripted Monologue Smartphone Speech Dataset | 786 Hours | Multilingual Language Data
- Dutch Speech-to-Text Transcribed Audio Dataset
- Dutch Unscripted Call Center Telephony speech dataset | 166 Hours | Speech Data
- E-commerce Return Call Text Dataset for Retail Analysis (United States, English)
- Earnings Call Transcripts & Investor Presentations Data | 26K+ Companies, Global incl. Europe & Asia | API, Markdown