README.md

metadata

license: mit
language:
  - en
  - ru
task_categories:
  - image-to-text
tags:
  - text-scene-detection
  - text-detection
  - yolo
size_categories:
  - 10K<n<100K

The text scene detection dataset is marked up manually in yolo format and contains 24 053 images, mostly with Latin and Russian text.