SignWriting Database

The SW database is composed of 7997 binary images divided into 103 classes, which were defined after a joint study with the Brazilian deaf community. For more details, please refer to [1,2]. Up to now, 30 people, deaf and non-deaf, contributed to the database. The next figure shows the 103 classes of the database

Some of the variability of the handwriting can be observed in the next figure

How to obtain access to the images

The SW Database may be used for non-commercial research provided you acknowledge the source of the image by citing the following paper in publications about your research:

  • D. Stiehl, L. S. Oliveira, C. Guimarães, A. S. Britto Jr, Towards a SignWriting Recognition System, International Conference on Document Analysis and Recognition (ICDAR 2015), Nancy, France, 2015. (pdf).

Related papers

[1] C. Guimarães, J. F. Guardezi, and S. Fernandes. Sign language writing acquisition – technology for a writing system. In 47th Hawaii International Conference on System Science, pages 120–129. IEEE, 2014.

[2] C. Guimarães, S. Fernandes, J. F. Guardezi, and L. S. Oliveira. Deaf culture and sign language writing system: a database for a new approach to writing system recognition technology. In 47th Hawaii International Conference on System Science, pages 3368–3377, 2014.

This database is licensed under a Creative Commons Attribution 4.0 International License.