Matěj Kocián
Matěj Kocián has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Ahmed Hamdi
University of Toulouse, IRIT, Toulouse, France
3 shared publications - Antoine Doucet
University of La Rochelle, La Rochelle, France
3 shared publications - Dimosthenis Karatzas
Computer Vision Center
3 shared publications - Jiri Matas
Center for Machine Perception, Czech Technical University, Prague, Czechia
3 shared publications - Matyáš Skalický
Rossum
3 shared publications - Michal Uřičář
Rossum
3 shared publications
Latest Publications
- 11 September 2023
- Published by Springer Nature
- p. 276-293
Overview of DocILE 2023: Document Information Localization and Extraction
- 11 September 2023
- Published by Springer Nature
- p. 276-293
Overview of DocILE 2023: Document Information Localization and Extraction
- 19 August 2023
- Published by Springer Nature
- p. 147-166
DocILE Benchmark for Document Information Localization and Extraction
- 19 August 2023
- Published by Springer Nature
- p. 147-166
DocILE Benchmark for Document Information Localization and Extraction
- Open Access
DocILE Benchmark for Document Information Localization and Extraction
This paper introduces the DocILE benchmark with the largest dataset of business documents for the tasks of Key Information Localization and Extraction and Line Item Recognition. It contains 6.7k annotated business documents, 100k synthetically generated documents, and nearly~1M unlabeled documents for unsupervised pre-training. The dataset has been built with knowledge of domain- and task-specific aspects, resulting in the following key features: (i) annotations in 55 classes, which surpasses the granularity of previously published key information extraction datasets by a large margin; (ii) Line Item Recognition represents a highly practical information extraction task, where key information has to be assigned to items in a table; (iii) documents come from numerous layouts and the test set includes zero- and few-shot cases as well as layouts commonly seen in the training set. The benchmark comes with several baselines, including RoBERTa, LayoutLMv3 and DETR-based Table Transformer; applied to both tasks of the DocILE benchmark, with results shared in this paper, offering a quick starting point for future work. The dataset, baselines and supplementary material are available at https://github.com/rossumai/docile.
Open AccessDocILE Benchmark for Document Information Localization and Extraction
This paper introduces the DocILE benchmark with the largest dataset of business documents for the tasks of Key Information Localization and Extraction and Line Item Recognition. It contains 6.7k annotated business documents, 100k synthetically generated documents, and nearly~1M unlabeled documents for unsupervised pre-training. The dataset has been built with knowledge of domain- and task-specific aspects, resulting in the following key features: (i) annotations in 55 classes, which surpasses the granularity of previously published key information extraction datasets by a large margin; (ii) Line Item Recognition represents a highly practical information extraction task, where key information has to be assigned to items in a table; (iii) documents come from numerous layouts and the test set includes zero- and few-shot cases as well as layouts commonly seen in the training set. The benchmark comes with several baselines, including RoBERTa, LayoutLMv3 and DETR-based Table Transformer; applied to both tasks of the DocILE benchmark, with results shared in this paper, offering a quick starting point for future work. The dataset, baselines and supplementary material are available at https://github.com/rossumai/docile.