Skip to content

Latest commit

 

History

History
70 lines (52 loc) · 3.24 KB

README.md

File metadata and controls

70 lines (52 loc) · 3.24 KB
Raw

hdBPMN - hand-drawn BPMN dataset

This repository contains the hand-drawn BPMN dataset published jointly with our CAiSE 2021 paper Sketch2BPMN: Automatic Recognition of Hand-drawn BPMN Models by Bernhard Schäfer, Han van der Aa, Henrik Leopold and Heiner Stuckenschmidt.

hdBPMN Example Image

Releases

Related Repositories

  • pybpmn-parser: contains a BPMN XML parser and a script for creating a dataset in the COCO format (COCO is a common format for computer vision datasets).
  • handwritten-diagram-datasets: hosts the generated COCO datasets among other diagram datasets and features a demo that shows how to parse and visualize the COCO annotations
  • bpmn-image-annotator: Annotation tool that has been developed to annotate hand-drawn images with ground-truth BPMN models.

Demo

Also check out our Sketch2BPMN demo where you can try our method with your own images and download the generated BPMN XML.

Project Organization

├── data
│   ├── annotations         <- ex00-ex10 .bpmn files.
│   ├── images              <- ex00-ex10 .png/.jpg images.
│   ├── words               <- ex00-ex10 Pascal VOC word annotation .xml files.
│   ├── exercises.pdf       <- modeling tasks ex00-ex10 that the hand-drawn models correspond to.
│   ├── writer_split.csv    <- maps each writerID to one of train/val/test split
└── README.md               <- Current file

License

The dataset is licensed under a Creative Commons Attribution 4.0 International License.

CC BY 4.0

Citing hdBPMN

@inproceedings{schaeferSketch2BPMN2021,
  title = {Sketch2BPMN: Automatic Recognition of Hand-drawn BPMN Models},
  author = {Sch{\"a}fer, Bernhard and {van der Aa}, Han and Leopold, Henrik and Stuckenschmidt, Heiner},
  booktitle = {Advanced Information Systems Engineering},
  year = {2021},
  publisher = {{Springer International Publishing}},
  language = {en},
  series = {Lecture {{Notes}} in {{Computer Science}}}
}
@inproceedings{schaeferDiagramNet2021,
  title={DiagramNet: Hand-drawn Diagram Recognition using Visual Arrow-relation Detection},
  author = {Sch{\"a}fer, Bernhard and Stuckenschmidt, Heiner},
  booktitle={2021 International Conference on Document Analysis and Recognition (ICDAR)},
  year={2021},
  month = sep,
  language = {en}
}