Datasets

Several datasets have been collected within this project. These datasets can freely be used for non-commercial usage with the condition of properly citing the corresponding publication. The list of these datasets is given below:

AHDB Dataset

This dataset has been collected in Qatar University and is essentially meant for Arabic Handwriting Recognition tasks is available free for non-commercial research. It can be downloaded by sending the signed REGISTRATION FORM to s_alali@qu.edu.qa. We accept the professor/supervisor to register for their students.

Related publication:

1. S. Al-Ma’adeed, D. Elliman, C.A. Higgins, A database for Arabic handwritten text recognition research, in 8th International Workshop on Frontiers in Handwriting Recognition, Niagara-on-the-Lake, Canada, 2002.
2. H. Hassen, S. Al-Ma’adeed, Arabic Handwriting Recognition using Sequential Minimal Optimization , in the 1st International Workshop on Arabic Script Analysis and Recognition (ASAR 2017), Loria Nancy, France.

ICDAR2011 - Arabic Writer Identification Contest Dataset

This dataset has been used in the Arabic Writer Identification Contest which has been organized in the scope of ICDAR2011. The data is available for download here.

Related publication:

A. Hassaïne, S. Al-Ma’adeed, J. Alja’am, A. Jaoua and A. Bouridane, The ICDAR2011 Arabic Writer Identification Contest, in 11th International Conference on Document Analysis and Recognition, Beijing, China. September 2011.

ICHFR2012 - Writer Identification Contest for Arabic Scripts Dataset

This dataset has been used in the Writer Identification Contest for Arabic Scripts which has been organized in the scope of ICHFR2012. The data is available for download here.

Related publication:

A. Hassaïne and S. Al Maadeed, ICFHR2012 Competition on Writer Identification – Challenge 2: Arabic Scripts, in 13th International Conference on Frontiers in Handwriting Recognition, p. 831-836, 2012.

QUWI Handwriting Dataset

This dataset has been collected within this project. It contains handwritings of over 1000 writers in both Arabic and English. This dataset can be used for both writer identification and handwriting recognition tasks. It can be downloaded by sending the signed REGISTRATION FORM to s_alali@qu.edu.qa. We accept the professor/supervisor to register for their students.

Related publication:

1. Al-Maadeed, Somaya, et al. “Novel geometric features for off-line writer identification.” Pattern Analysis and Applications 19.3 (2016): 699-708.
2. Al Maadeed, Somaya, and Abdelaali Hassaine. “Automatic prediction of age, gender, and nationality in offline handwriting.” EURASIP Journal on Image and Video Processing 2014.1 (2014): 1-10.
3. S. Al-Maadeed, W. Ayouby, A. Hassaïne, J. Alja’am, QUWI: An Arabic and English Handwriting Dataset for Offline Writer Identification, in 13th International Conference on Frontiers in Handwriting Recognition, p. 742-747, 2012.
4. Hassaïne, Abdelaali, et al. “The ICDAR2011 Arabic writer identification contest.” Document Analysis and Recognition (ICDAR), 2011 International Conference on. IEEE, 2011.

© 2016 Qatar University
Hosted by Qatar University