The Impact of Visual Similarities of Arabic-Like Scripts Regarding Learning in an OCR System
作者： Riaz Ahmad , Saeeda Naz , M. Zeshan Afzal , S. Faisal Rashid , Marcus Liwicki
关键词: Arabic script 、 Artificial intelligence 、 Scripting language 、 Natural language processing 、 Computer science 、 Optical character recognition 、 Urdu 、 Transfer of learning 、 Arabic 、 Pashto 、 Persian
摘要: Many languages use Arabic script for written communication either in basic or augmented form. These include Urdu, Pashto, Persian, etc. As the primary characters are shared among all these languages, it is possible to take advantage of visual similarities Optical Character Recognition (OCR). OCR models optimized individual have been proposed. However, best our knowledge, there no attempt develop a single system more than one language. The contributions presented work are: First, investigates effect on recognition accuracy when different combined (A pioneering study). Second, introduces publicly available synthetic datasets and Pashto experimental purposes. Third, this paper provides statistical analysis as clues transfer learning concerning systems Arabic, languages.
sci-hub.se 参考文章(18)
Saeeda Naz, Arif I. Umar, Riaz Ahmad, Saad B. Ahmed, Syed H. Shirazi, Muhammad I. Razzak, Urdu Nasta'liq text recognition system based on multi-dimensional recurrent neural network and statistical features Neural Computing and Applications. ,vol. 28, pp. 219- 231 ,(2017) , 10.1007/S00521-015-2051-4
V. I. Levenshtein, Binary codes capable of correcting deletions, insertions, and reversals Soviet physics. Doklady. ,vol. 10, pp. 707- 710 ,(1966)
Mohammad Tanvir Parvez, Sabri A. Mahmoud, Offline arabic handwritten text recognition: A Survey ACM Computing Surveys. ,vol. 45, pp. 23- ,(2013) , 10.1145/2431211.2431222
Mohammad Reza Yousefi, Mohammad Reza Soheili, Thomas M. Breuel, Didier Stricker, A comparison of 1D and 2D LSTM architectures for the recognition of handwritten Arabic document recognition and retrieval. ,vol. 9402, ,(2015) , 10.1117/12.2075930
Sabri A. Mahmoud, Irfan Ahmad, Wasfi G. Al-Khatib, Mohammad Alshayeb, Mohammad Tanvir Parvez, Volker Märgner, Gernot A. Fink, KHATT: An open Arabic offline handwritten text database Pattern Recognition. ,vol. 47, pp. 1096- 1112 ,(2014) , 10.1016/J.PATCOG.2013.08.009
Nazly Sabbour, Faisal Shafait, A segmentation-free approach to Arabic and Urdu OCR document recognition and retrieval. ,vol. 8658, pp. 1- 12 ,(2013) , 10.1117/12.2003731
Sheikh Faisal Rashid, Marc-Peter Schambach, Jörg Rottland, Stephan von der Nüll, Low resolution Arabic recognition with multidimensional recurrent neural networks Proceedings of the 4th International Workshop on Multilingual OCR. pp. 6- ,(2013) , 10.1145/2505377.2505385
Sabri A. Mahmoud, Irfan Ahmad, Mohammad Alshayeb, Wasfi G. Al-Khatib, Mohammad Tanvir Parvez, Gernot A. Fink, Volker Margner, Haikal El Abed, KHATT: Arabic Offline Handwritten Text Database international conference on frontiers in handwriting recognition. pp. 449- 454 ,(2012) , 10.1109/ICFHR.2012.224
A. Graves, M. Liwicki, S. Fernandez, R. Bertolami, H. Bunke, J. Schmidhuber, A Novel Connectionist System for Unconstrained Handwriting Recognition IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 31, pp. 855- 868 ,(2009) , 10.1109/TPAMI.2008.137
Riaz Ahmad, Muhammad Zeshan Afzal, Sheikh Faisal Rashid, Marcus Liwicki, Thomas Breuel, Scale and rotation invariant OCR for Pashto cursive script using MDLSTM network international conference on document analysis and recognition. pp. 1101- 1105 ,(2015) , 10.1109/ICDAR.2015.7333931