Datasets:

Netra-Lab
/

khob-eval

license: mit
language:
  - km
task_categories:
  - image-to-text
tags:
  - ocr
  - khmer-ocr
  - benchmark
pretty_name: KHOB (325 lines)
size_categories:
  - n<1K

KHOB, 325 text lines

325 Khmer text-line images with their transcriptions, cropped from the Khmer OCR Benchmark (KHOB) by EKYC Solutions: scanned Cambodian government documents with clean backgrounds and compression artefacts. One image is one line.

This is the KHOB test set in the Netra OCR results, where the recognizer scores 0.96% CER with beam search and 1.29% greedy.

Column
`image`	the line crop
`label`	its text
`file_name`	the original crop's name: task, document and line number

Use

from datasets import load_dataset
ds = load_dataset("Netra-Lab/khob-eval", split="train")

With the Netra OCR repository, the khob preset scores any model on it:

bash training/evaluate.sh published --test khob

License

MIT, as the original KHOB dataset. Credit: EKYC Solutions, khmer-ocr-benchmark-dataset.