Cite Article

Trace Transform Feature Learning for Offline Jawi Handwritten Recognition

Choose citation format

BibTeX

@article{IJASEIT10113,
   author = {Anton Heryanto Hasan and Khairuddin Omar and Muhammad Faidzul Nasrudin},
   title = {Trace Transform Feature Learning for Offline Jawi Handwritten Recognition},
   journal = {International Journal on Advanced Science, Engineering and Information Technology},
   volume = {11},
   number = {1},
   year = {2021},
   pages = {37--42},
   keywords = {Jawi; handwritten recognition; sub-word; end-to-end learning; features learning; trace transform.},
   abstract = {

Offline Jawi handwritten recognition is very important to allow efficient archiving and retrieving the original documents and increase the availability of the content. It is challenging task and still considered an open problem because the state-of-the-art recognizer performance is considered sub-par. The tradition trace Transform features extractor has potential, however the complexity of parameters tuning in feature engineered approach combine with independent non-learnable sub-word classifier produce sub-par Jawi sub-word recognition accuracy. The proposed trace Transform feature learning address the features extraction complexity by automatically discovers the features according to data. The features extractor and classifier trained end-to-end from raw input data to target class to find the optimum parameters. The trace transform process defined as layer similar with convolution process in Convolution Neural Network. This approach improves data representation and produce better Jawi handwritten recognition performance. trace Transform feature learning are more robust to Affine Transformations compared to the state-of-the-arts Convolution Neural Networks feature learning because its data representation invariant to rotation, slanting and skewing. This proposed feature learning performance evaluated with its performance in sub-word recognition performance using Jawi dataset. In this paper only single layer of trace transform feature learning compare with traditional trace transform feature and Convolution Neural network as the state-of-the-art feature learning. The performances are significantly better compared to traditional trace transform feature and able to compete with convolution neural network in single layer, three layers and comparable with eight layers.

},    issn = {2088-5334},    publisher = {INSIGHT - Indonesian Society for Knowledge and Human Development},    url = {http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=10113},    doi = {10.18517/ijaseit.11.1.10113} }

EndNote

%A Hasan, Anton Heryanto
%A Omar, Khairuddin
%A Nasrudin, Muhammad Faidzul
%D 2021
%T Trace Transform Feature Learning for Offline Jawi Handwritten Recognition
%B 2021
%9 Jawi; handwritten recognition; sub-word; end-to-end learning; features learning; trace transform.
%! Trace Transform Feature Learning for Offline Jawi Handwritten Recognition
%K Jawi; handwritten recognition; sub-word; end-to-end learning; features learning; trace transform.
%X 

Offline Jawi handwritten recognition is very important to allow efficient archiving and retrieving the original documents and increase the availability of the content. It is challenging task and still considered an open problem because the state-of-the-art recognizer performance is considered sub-par. The tradition trace Transform features extractor has potential, however the complexity of parameters tuning in feature engineered approach combine with independent non-learnable sub-word classifier produce sub-par Jawi sub-word recognition accuracy. The proposed trace Transform feature learning address the features extraction complexity by automatically discovers the features according to data. The features extractor and classifier trained end-to-end from raw input data to target class to find the optimum parameters. The trace transform process defined as layer similar with convolution process in Convolution Neural Network. This approach improves data representation and produce better Jawi handwritten recognition performance. trace Transform feature learning are more robust to Affine Transformations compared to the state-of-the-arts Convolution Neural Networks feature learning because its data representation invariant to rotation, slanting and skewing. This proposed feature learning performance evaluated with its performance in sub-word recognition performance using Jawi dataset. In this paper only single layer of trace transform feature learning compare with traditional trace transform feature and Convolution Neural network as the state-of-the-art feature learning. The performances are significantly better compared to traditional trace transform feature and able to compete with convolution neural network in single layer, three layers and comparable with eight layers.

%U http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=10113 %R doi:10.18517/ijaseit.11.1.10113 %J International Journal on Advanced Science, Engineering and Information Technology %V 11 %N 1 %@ 2088-5334

IEEE

Anton Heryanto Hasan,Khairuddin Omar and Muhammad Faidzul Nasrudin,"Trace Transform Feature Learning for Offline Jawi Handwritten Recognition," International Journal on Advanced Science, Engineering and Information Technology, vol. 11, no. 1, pp. 37-42, 2021. [Online]. Available: http://dx.doi.org/10.18517/ijaseit.11.1.10113.

RefMan/ProCite (RIS)

TY  - JOUR
AU  - Hasan, Anton Heryanto
AU  - Omar, Khairuddin
AU  - Nasrudin, Muhammad Faidzul
PY  - 2021
TI  - Trace Transform Feature Learning for Offline Jawi Handwritten Recognition
JF  - International Journal on Advanced Science, Engineering and Information Technology; Vol. 11 (2021) No. 1
Y2  - 2021
SP  - 37
EP  - 42
SN  - 2088-5334
PB  - INSIGHT - Indonesian Society for Knowledge and Human Development
KW  - Jawi; handwritten recognition; sub-word; end-to-end learning; features learning; trace transform.
N2  - 

Offline Jawi handwritten recognition is very important to allow efficient archiving and retrieving the original documents and increase the availability of the content. It is challenging task and still considered an open problem because the state-of-the-art recognizer performance is considered sub-par. The tradition trace Transform features extractor has potential, however the complexity of parameters tuning in feature engineered approach combine with independent non-learnable sub-word classifier produce sub-par Jawi sub-word recognition accuracy. The proposed trace Transform feature learning address the features extraction complexity by automatically discovers the features according to data. The features extractor and classifier trained end-to-end from raw input data to target class to find the optimum parameters. The trace transform process defined as layer similar with convolution process in Convolution Neural Network. This approach improves data representation and produce better Jawi handwritten recognition performance. trace Transform feature learning are more robust to Affine Transformations compared to the state-of-the-arts Convolution Neural Networks feature learning because its data representation invariant to rotation, slanting and skewing. This proposed feature learning performance evaluated with its performance in sub-word recognition performance using Jawi dataset. In this paper only single layer of trace transform feature learning compare with traditional trace transform feature and Convolution Neural network as the state-of-the-art feature learning. The performances are significantly better compared to traditional trace transform feature and able to compete with convolution neural network in single layer, three layers and comparable with eight layers.

UR - http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=10113 DO - 10.18517/ijaseit.11.1.10113

RefWorks

RT Journal Article
ID 10113
A1 Hasan, Anton Heryanto
A1 Omar, Khairuddin
A1 Nasrudin, Muhammad Faidzul
T1 Trace Transform Feature Learning for Offline Jawi Handwritten Recognition
JF International Journal on Advanced Science, Engineering and Information Technology
VO 11
IS 1
YR 2021
SP 37
OP 42
SN 2088-5334
PB INSIGHT - Indonesian Society for Knowledge and Human Development
K1 Jawi; handwritten recognition; sub-word; end-to-end learning; features learning; trace transform.
AB 

Offline Jawi handwritten recognition is very important to allow efficient archiving and retrieving the original documents and increase the availability of the content. It is challenging task and still considered an open problem because the state-of-the-art recognizer performance is considered sub-par. The tradition trace Transform features extractor has potential, however the complexity of parameters tuning in feature engineered approach combine with independent non-learnable sub-word classifier produce sub-par Jawi sub-word recognition accuracy. The proposed trace Transform feature learning address the features extraction complexity by automatically discovers the features according to data. The features extractor and classifier trained end-to-end from raw input data to target class to find the optimum parameters. The trace transform process defined as layer similar with convolution process in Convolution Neural Network. This approach improves data representation and produce better Jawi handwritten recognition performance. trace Transform feature learning are more robust to Affine Transformations compared to the state-of-the-arts Convolution Neural Networks feature learning because its data representation invariant to rotation, slanting and skewing. This proposed feature learning performance evaluated with its performance in sub-word recognition performance using Jawi dataset. In this paper only single layer of trace transform feature learning compare with traditional trace transform feature and Convolution Neural network as the state-of-the-art feature learning. The performances are significantly better compared to traditional trace transform feature and able to compete with convolution neural network in single layer, three layers and comparable with eight layers.

LK http://ijaseit.insightsociety.org/index.php?option=com_content&view=article&id=9&Itemid=1&article_id=10113 DO - 10.18517/ijaseit.11.1.10113