A deep learning based Arabic script recognition system : benchmark on KHAT

Joint Authors

Naz, Saidah
Rashid, Shaykh
Liwicki, Marcus
Dengel, Andreas
Ahmad, Riaz
Afzal, Muhammad

Source

The International Arab Journal of Information Technology

Issue

Vol. 17, Issue 3 (31 May. 2020), pp.299-305, 7 p.

Publisher

Zarqa University Deanship of Scientific Research

Publication Date

2020-05-31

Country of Publication

Jordan

No. of Pages

7

Main Subjects

Information Technology and Computer Science

Abstract EN

This paper presents a deep learning benchmark on a complex dataset known as KFUPM Handwritten Arabic TexT (KHATT).

The KHATT data-set consists of complex patterns of handwritten Arabic text-lines.

This paper contributes mainly in three aspects i.e., (1) pre-processing, (2) deep learning based approach, and (3) data-augmentation.

The pre-processing step includes pruning of white extra spaces plus de-skewing the skewed text-lines.

We deploy a deep learning approach based on Multi-Dimensional Long Short-Term Memory (MDLSTM) networks and Connectionist Temporal Classification (CTC).

The MDLSTM has the advantage of scanning the Arabic text-lines in all directions (horizontal and vertical) to cover dots, diacritics, strokes and fine inflammation.

The data-augmentation with a deep learning approach proves to achieve better and promising improvement in results by gaining 80.02% Character Recognition (CR) over 75.08% as baseline.

American Psychological Association (APA)

Ahmad, Riaz& Naz, Saidah& Afzal, Muhammad& Rashid, Shaykh& Liwicki, Marcus& Dengel, Andreas. 2020. A deep learning based Arabic script recognition system : benchmark on KHAT. The International Arab Journal of Information Technology،Vol. 17, no. 3, pp.299-305.
https://search.emarefa.net/detail/BIM-962332

Modern Language Association (MLA)

Ahmad, Riaz…[et al.]. A deep learning based Arabic script recognition system : benchmark on KHAT. The International Arab Journal of Information Technology Vol. 17, no. 3 (May. 2020), pp.299-305.
https://search.emarefa.net/detail/BIM-962332

American Medical Association (AMA)

Ahmad, Riaz& Naz, Saidah& Afzal, Muhammad& Rashid, Shaykh& Liwicki, Marcus& Dengel, Andreas. A deep learning based Arabic script recognition system : benchmark on KHAT. The International Arab Journal of Information Technology. 2020. Vol. 17, no. 3, pp.299-305.
https://search.emarefa.net/detail/BIM-962332

Data Type

Journal Articles

Language

English

Notes

Includes bibliographical references : p. 303-304

Record ID

BIM-962332