Skip to content

Dataset · GreenNode

Synthetic document pages for OCR

GreenNode/GreenOCR-Dataset

Generated pages in Vietnamese and English with block annotations for training document parsing.

What's inside

Contains 1,000 to 10,000 synthetic document pages in Vietnamese and English, along with their augmented versions simulating scans, smartphone photos, and aged paper. Each page includes precise annotations: text block polygons, their content, and ready targets in LightOnOCR-3 format.

What you can do with it

  • Training or fine-tuning OCR and document structure parsing models
  • Fine-tuning vision-language models (VLMs) for coordinate-grounded text extraction
  • Testing recognition robustness against scan artifacts and poor photo quality
Tasks
Image → text, Image text → text
Languages
Vietnamese, English
Size
1K–10K rows
Downloads this month
49.4K
Likes
1
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("GreenNode/GreenOCR-Dataset")