datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
text_recognition_en_zh_clean
Dataset Card for "text_recognition_en_zh_clean"
More Information needed
MJSynth_text_recognition
Dataset Card for "MJSynth_text_recognition"
This is the MJSynth dataset for text recognition on document images, synthetically generated, covering 90K English words.
It includes training, validation and test splits.
Source of the dataset: https://www.robots.ox.ac.uk/~vgg/data/text/
Use dataset streaming functionality to try out the dataset quickly without downloading the entire dataset (refer: https://huggingface.co/docs/datasets/stream)
Citation details provided on the source… See the full description on the dataset page: https://huggingface.co/datasets/priyank-m/MJSynth_text_recognition.trdg_random_en_zh_text_recognition
Dataset Card for "trdg_random_en_zh_text_recognition"
This synthetic dataset was generated using the TextRecognitionDataGenerator(TRDG) open source repo:
https://github.com/Belval/TextRecognitionDataGenerator
It contains images of text with random characters from Engilsh(en) and Chinese(zh) languages.
Reference to the documentation provided by the TRDG repo:
https://textrecognitiondatagenerator.readthedocs.io/en/latest/index.html
text_recognition_en_zh
Dataset Card for "text_recognition_en_zh"
More Information needed
synth-text-recognition
Dataset Card for "Synth-Text Recognition"
This is the dataset for text recognition on document images, synthetically generated, covering 90K English words.
It includes training, validation and test splits.
text-recognition
Dataset Card for "text-recognition"
More Information needed
chinese_text_recognitionSource of data: https://github.com/FudanVI/benchmarking-chinese-text-recognition
synth-text-recognition-multilines-cs
Czech Synthetic Multiline Text Recognition Dataset
A large-scale synthetic dataset for Czech multiline text recognition, containing 100,000 text images with corresponding transcriptions. Created using SynthTiger.
Dataset Description
This dataset consists of synthetically generated images of Czech text with multiple lines per image, designed for training optical character recognition (OCR) models that can handle complex multiline text layouts. Each image contains 3 lines… See the full description on the dataset page: https://huggingface.co/datasets/Empatixx/synth-text-recognition-multilines-cs.synth-text-recognition-cs
Czech Synthetic Text Recognition Dataset
A large-scale synthetic dataset for Czech text recognition, containing 454,820 text images with corresponding transcriptions. Created using SynthTiger.
Dataset Description
This dataset consists of synthetically generated images of Czech text, designed for training optical character recognition (OCR) models. Each image contains a single word or short phrase rendered with various visual effects to simulate real-world text appearance.… See the full description on the dataset page: https://huggingface.co/datasets/Empatixx/synth-text-recognition-cs.IAM_words_text_recognitiontrdg_wikipedia_en_text_recognition
Dataset Card for "trdg_wikipedia_en_zh_text_recognition"
This synthetic dataset was generated using the TextRecognitionDataGenerator(TRDG) open source repo:
https://github.com/Belval/TextRecognitionDataGenerator
It contains synthetic images of text randomly sampled from Engilsh(en) Wikipedia pages.
Reference to the documentation provided by the TRDG repo:
https://textrecognitiondatagenerator.readthedocs.io/en/latest/index.html
text_recognition_en_zh_250k
Dataset Card for "text_recognition_en_zh_250k"
More Information needed
trdg_random_single_words_en_text_recognition
Dataset Card for "trdg_random_single_words_en_text_recognition"
More Information needed
license-plate-text-recognition-full
Dataset Card for "license-plate-text-recognition-full"
Background Information
This dataset is generated from keremberke/license-plate-object-detection dataset. What we have done is:
Get the Bounding Boxes for each plate in an image,
Crop the image to make the plate only visible,
Run it through the microsoft/trocr-large-printed model to extract the written information.
Structure of the Dataset
It has the same structure as the… See the full description on the dataset page: https://huggingface.co/datasets/sonnetechnology/license-plate-text-recognition-full.text_recognition_en_zh_small_250k
Dataset Card for "text_recognition_en_zh_small_250k"
More Information needed
Arabic-Handwritten-Text-Recognition-Dataset
Dataset Description
This dataset is a re-uploaded version of the Muharaf dataset.
The original dataset was created by Mehreen Saeed et al. and released for research purposes.
This repository is intended for easier access and experimentation via Hugging Face.
|
How to use
from datasets import load_dataset
ds = load_dataset("amjad-awad/Arabic-Handwritten-Text-Recognition-Dataset")
print(ds["train"][0]["image"])
Attribution
All credit for creating and… See the full description on the dataset page: https://huggingface.co/datasets/amjad-awad/Arabic-Handwritten-Text-Recognition-Dataset.iam_sroie_text_recognitiontrdg_dict_random_words_en_text_recognition
Dataset Card for "trdg_random_words_en_text_recognition"
More Information needed
handwritten-text-recognition-bongabdo
Dataset Card for Bongabdo
Dataset Summary
Bongabdo is a curated dataset of full-page Bangla (Bengali) handwritten text, intended for use in offline handwriting recognition tasks using modern neural architectures. It includes high-resolution scanned images of handwritten Bangla scripts, transcriptions, and rich per-document metadata. The data has been contributed by people of diverse age groups, occupations, and genders, making it well-suited for training robust… See the full description on the dataset page: https://huggingface.co/datasets/deepcopy/handwritten-text-recognition-bongabdo.thvl_text_recognition
Dataset Card for "thvl_text_recognition"
More Information needed
balanced_SROIE_CHINESE_IAM_text_recognitionchinese_text_recognitionSource of data: https://github.com/FudanVI/benchmarking-chinese-text-recognition
word_based_IAM_SROIE_text_recognition
Dataset Card for "word_based_IAM_SROIE_text_recognition"
More Information needed
khmer-text-recognitionamharic-text-recognitionSROIE_2019_text_recognition-other-cols-5Subset (+ some renamings and data processing) of https://huggingface.co/datasets/priyank-m/SROIE_2019_text_recognition
khmer-text-recognitiontext_recognition_TextVQAEvaluate with VQA Accuracy: Note! A question might have various correct answers! So if more than 3 answers are the same, then scores 1; otherwise, scores with proportion.
(here is a evaluation function that you can use)
def vqa_accuracy(predictions, ground_truths_list):
accuracies = []
for pred, ground_truths in zip(predictions, ground_truths_list):
answer_count = collections.Counter(ground_truths)
accuracy = min(1.0, answer_count[pred] / 3.0)… See the full description on the dataset page: https://huggingface.co/datasets/Holmes377/text_recognition_TextVQA.