datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ICDAR2019-SROIE
ICDAR2019's Scanned Receipts OCR and Information Extraction (SROIE)
The ICDAR2019 SROIE dataset was originally published by Huang et al. for the
15th International Conference on Document Analysis and Recognition (ICDAR2019)
Robust Reading Challenge on Scanned Receipts OCR and Information Extraction
(SROIE).
This work presents an extension of the original ICDAR2019 SROIE dataset, including 14
receipt annotations missing from the original Task 3 test dataset, in a format
integrated… See the full description on the dataset page: https://huggingface.co/datasets/jsdnrs/ICDAR2019-SROIE.ICDAR2015icdar2021-historical-document-dating
ICDAR 2021 Historical Document Classification — Task 2 (Dating)
13,810 manuscript page images labelled with the period in which they were produced.
Images come from e-codices, the virtual manuscript library
of Switzerland.
Split
Images
Date range
Median span
Dated to a single year
train
11,294
800–1899
45 years
1,409
test
2,516
800–1921
49 years
264
The label is an interval, not a year
Palaeographers date a manuscript to a range, and the width of… See the full description on the dataset page: https://huggingface.co/datasets/biglam/icdar2021-historical-document-dating.icdar_disco
ICDAR_mini Dataset
A balanced mini subset of the ICDAR (International Conference on Document Analysis and Recognition) dataset with 50 samples per language. Includes actual document images and ground truth OCR text.
Dataset Details
Total Samples: 500
Total Images: 500
Languages: 10
Arabic (50 samples)
Bangla (50 samples)
Chinese (50 samples)
Hindi (50 samples)
Japanese (50 samples)
Korean (50 samples)
Latin (50 samples)
Mixed (50 samples)
None (50 samples)… See the full description on the dataset page: https://huggingface.co/datasets/kenza-ily/icdar_disco.ICDAR2015_OCR
META
https://github.com/open-mmlab/mmocr/blob/main/dataset_zoo/icdar2015/metafile.yml
Name: 'Incidental Scene Text IC15'
Paper:
Title: ICDAR 2015 Competition on Robust Reading
URL: https://rrc.cvc.uab.es/files/short_rrc_2015.pdf
Venue: ICDAR
Year: '2015'
BibTeX: '@inproceedings{karatzas2015icdar,
title={ICDAR 2015 competition on robust reading},
author={Karatzas, Dimosthenis and Gomez-Bigorda, Lluis and Nicolaou, Anguelos and Ghosh, Suman and Bagdanov, Andrew and… See the full description on the dataset page: https://huggingface.co/datasets/MiXaiLL76/ICDAR2015_OCR.icdar-htr-competition-2015
ICDAR 2015 HTRtS Competition Dataset
Dataset Description
The ICDAR 2015 Competition HTRtS dataset was used for the "Handwritten Text Recognition on the tranScriptorium Dataset" competition held at the International Conference on Document Analysis and Recognition (ICDAR) in 2015.
The dataset images are drawn from the English "Bentham collection", originally used in the European-funded TRAN SCRIPTORIUM project.
The data features significant difficulties, including:… See the full description on the dataset page: https://huggingface.co/datasets/SSamDav/icdar-htr-competition-2015.icdar2021-mapseg
ICDAR 2021 MapSeg — historical map segmentation
Map sheets from nine atlases of the City of Paris produced between 1894 and 1937 by the city's Map
Service (Service du plan), annotated for three segmentation tasks. Built for the ICDAR 2021
Competition on Historical Map Segmentation and deposited on Zenodo
(10.5281/zenodo.4817662).
This repository is a format conversion of that deposit — same images, same ground truth, repackaged so
it loads with datasets instead of a 1.9 GB… See the full description on the dataset page: https://huggingface.co/datasets/biglam/icdar2021-mapseg.icdar-2019-competition-cbad
ICDAR 2019 Competition on Baseline Detection (cBAD) Dataset
Dataset Description
The ICDAR 2019 Competition on Baseline Detection (cBAD) dataset was developed for the eponymous competition focused on Baseline Detection in historical documents. Baseline detection is a critical step for segmenting handwritten lines before Handwritten Text Recognition (HTR).
This is a newly created, real-world dataset consisting of 3,021 annotated document page images collected from seven… See the full description on the dataset page: https://huggingface.co/datasets/SSamDav/icdar-2019-competition-cbad.ICDAR_2025_Handwritten_Notes_Understanding_ChallengeICDAR-2011
ICDAR 2011 Signature Verification Competition (SigComp2011)
Description
The collection contains simultaneously acquired online and offline samples.
The collection contains offline and online signature samples.
The offline dataset comprises PNG images, scanned at 400 dpi, RGB color.
The online dataset comprises ascii files with the format: X, Y, Z (per line).
Marcus Liwicki, Michael Blumenstein, Elisa van den Heuvel, Charles E.H. Berger, Reinoud D. Stoel, Bryan… See the full description on the dataset page: https://huggingface.co/datasets/1aurent/ICDAR-2011.ICDAR-2024-SAM_Track_1ICDAR2013_OCR
META
https://github.com/open-mmlab/mmocr/blob/main/dataset_zoo/icdar2013/metafile.yml
Name: 'Incidental Scene Text IC13'
Paper:
Title: ICDAR 2013 Robust Reading Competition
URL: https://www.imlab.jp/publication_data/1352/icdar_competition_report.pdf
Venue: ICDAR
Year: '2013'
BibTeX:'@inproceedings{karatzas2013icdar,
title={ICDAR 2013 robust reading competition},
author={Karatzas, Dimosthenis and Shafait, Faisal and Uchida, Seiichi and Iwamura, Masakazu and i Bigorda… See the full description on the dataset page: https://huggingface.co/datasets/MiXaiLL76/ICDAR2013_OCR.ICDAR2019_cTDaR_TRACKB_resized
Dataset Card for ICDAR2019-cTDaR-TRACKB
This dataset is a resized version of the original cndplab-founder/ICDAR2019_cTDaR, merged with with its supplement cndplab-founder/ICDAR2019_cTDaR_dataset_supplement.
You can easily and quickly load it:
dataset = load_dataset("dvgodoy/ICDAR2019_cTDaR_TRACKB_resized")
DatasetDict({
train: Dataset({
features: ['image', 'width', 'height', 'category', 'label', 'bboxes_table', 'bboxes_cell'],
num_rows: 1200
})
test:… See the full description on the dataset page: https://huggingface.co/datasets/dvgodoy/ICDAR2019_cTDaR_TRACKB_resized.ICDAR-2024-HWD-Page-LevelICDAR2019-SROIE
ICDAR2019's Scanned Receipts OCR and Information Extraction (SROIE)
The ICDAR2019 SROIE dataset was originally published by Huang et al. for the
15th International Conference on Document Analysis and Recognition (ICDAR2019)
Robust Reading Challenge on Scanned Receipts OCR and Information Extraction
(SROIE).
This work presents an extension of the original ICDAR2019 SROIE dataset, including 14
receipt annotations missing from the original Task 3 test dataset, in a format… See the full description on the dataset page: https://huggingface.co/datasets/vishu12121/ICDAR2019-SROIE.icdar_sroieICDAR2013ICDAR2019_cTDaR_TRACKA_resized
Dataset Card for ICDAR2019-cTDaR-TRACKA
This dataset is a resized version of the original cndplab-founder/ICDAR2019_cTDaR.
You can easily and quickly load it:
dataset = load_dataset("dvgodoy/ICDAR2019_cTDaR_TRACKA_resized")
DatasetDict({
train: Dataset({
features: ['image', 'width', 'height', 'category', 'label', 'bboxes'],
num_rows: 1200
})
test: Dataset({
features: ['image', 'width', 'height', 'category', 'label', 'bboxes'],
num_rows:… See the full description on the dataset page: https://huggingface.co/datasets/dvgodoy/ICDAR2019_cTDaR_TRACKA_resized.OCR_ICDAR_English_HandwritingICDAR-2011
ICDAR 2011 Signature Verification Competition (SigComp2011)
Description
The collection contains simultaneously acquired online and offline samples.
The collection contains offline and online signature samples.
The offline dataset comprises PNG images, scanned at 400 dpi, RGB color.
The online dataset comprises ascii files with the format: X, Y, Z (per line).
Marcus Liwicki, Michael Blumenstein, Elisa van den Heuvel, Charles E.H. Berger, Reinoud D. Stoel, Bryan Found… See the full description on the dataset page: https://huggingface.co/datasets/Guvav0051/ICDAR-2011.icdar2013_train_cleaned
icdar2013_train_cleaned
The icdar2013_train family of the ElliotVL supervised-fine-tuning pool, after VLM cleaning.
images
227
QA turns
846
answers rewritten by the cleaning pass
88
QA created by the cleaning pass (new_qa)
661 (78.1%)
shards
1
How this was cleaned
A vision-language model read each image together with its QA and judged the item. The pass is
not a filter that only removes rows — it rewrites answers it finds wrong but… See the full description on the dataset page: https://huggingface.co/datasets/Elliot-Data/icdar2013_train_cleaned.icdar2015_train_cleaned
icdar2015_train_cleaned
The icdar2015_train family of the ElliotVL supervised-fine-tuning pool, after VLM cleaning.
images
968
QA turns
4,498
answers rewritten by the cleaning pass
874
QA created by the cleaning pass (new_qa)
3,531 (78.5%)
shards
1
How this was cleaned
A vision-language model read each image together with its QA and judged the item. The pass is
not a filter that only removes rows — it rewrites answers it finds wrong but… See the full description on the dataset page: https://huggingface.co/datasets/Elliot-Data/icdar2015_train_cleaned.icdar-2015-weatherICDAR2013_OCR
META
https://github.com/open-mmlab/mmocr/blob/main/dataset_zoo/icdar2013/metafile.yml
Name: 'Incidental Scene Text IC13'
Paper:
Title: ICDAR 2013 Robust Reading Competition
URL: https://www.imlab.jp/publication_data/1352/icdar_competition_report.pdf
Venue: ICDAR
Year: '2013'
BibTeX:'@inproceedings{karatzas2013icdar,
title={ICDAR 2013 robust reading competition},
author={Karatzas, Dimosthenis and Shafait, Faisal and Uchida, Seiichi and Iwamura, Masakazu and i Bigorda… See the full description on the dataset page: https://huggingface.co/datasets/Ashu44268/ICDAR2013_OCR.icdar2023vqabd-small-tables-val
Dataset Card for "icdar2023vqabd-small-tables-val"
More Information needed
icdar2015_train_cleaned
icdar2015_train_cleaned
The icdar2015_train family of the ElliotVL supervised-fine-tuning pool, after VLM cleaning.
images
968
QA turns
4,498
answers rewritten by the cleaning pass
874
QA created by the cleaning pass (new_qa)
3,531 (78.5%)
shards
1
How this was cleaned
A vision-language model read each image together with its QA and judged the item. The pass is
not a filter that only removes rows — it rewrites answers it finds wrong but… See the full description on the dataset page: https://huggingface.co/datasets/elliot-mllm/icdar2015_train_cleaned.icdar2013_train_cleaned
icdar2013_train_cleaned
The icdar2013_train family of the ElliotVL supervised-fine-tuning pool, after VLM cleaning.
images
227
QA turns
846
answers rewritten by the cleaning pass
88
QA created by the cleaning pass (new_qa)
661 (78.1%)
shards
1
How this was cleaned
A vision-language model read each image together with its QA and judged the item. The pass is
not a filter that only removes rows — it rewrites answers it finds wrong but… See the full description on the dataset page: https://huggingface.co/datasets/elliot-mllm/icdar2013_train_cleaned.ICDAR2019_cTDaR_TRACKAICDAR2019
Dataset Card for "ICDAR2019"
More Information needed
OCR_ICDAR_English_Handwriting