HMM Based Keyword Spotting System in Printed/Handwritten Arabic/Latin Documents with Identification Stage

Hybrid HMM/BLSTM system for multi-script keyword spotting in printed and handwritten documents with identification stage

Neural Computing and Applications ◽

10.1007/s00521-019-04429-w ◽

2019 ◽

Vol 32 (13) ◽

pp. 9201-9215

Author(s):

Ahmed Cheikhrouhou ◽

Yousri Kessentini ◽

Slim Kanoun

Keyword(s):

Keyword Spotting ◽

Handwritten Documents ◽

Identification Stage

Download Full-text

Model Fungsi Transfer Input Ganda untuk Pemodelan Jakarta Islamic Index

Xplore Journal of Statistics ◽

10.29244/xplore.v7i3.149 ◽

2019 ◽

Vol 7 (3) ◽

Author(s):

Nur Laela Fitriani ◽

Pika Silvianti ◽

Rahma Anisa

Keyword(s):

Exchange Rate ◽

Transfer Function ◽

Multivariate Time Series ◽

Model Identification ◽

Transfer Function Model ◽

Function Model ◽

Jones Index ◽

Multiple Input ◽

Identification Stage ◽

Stage Estimation

Transfer function model with multiple input is a multivariate time series forecasting model that combines several characteristics of ARIMA models by utilizing some regression analysis properties. This model is used to determine the effect of output series towards input series so that the model can be used to analyze the factors that affect the Jakarta Islamic Index (JII). The USD exchange rate against rupiah and Dow Jones Index (DJI) were used as input series. The transfer function model was constructed through several stages: model identification stage, estimation of transfer function model, and model diagnostic test. Based on the transfer function model, the JII was influenced by JII at the period of one and two days before. JII was also affected by the USD exchange rate against rupiah at the same period and at one and two days before. In addition, the JII was influenced by DJI at the same period and also at period of one until five days ago. The Mean Absolute Prencentage Error (MAPE) value of forecasting result was 0.70% and the correlation between actual and forecast data was 0.77. This shows that the model was well performed for forecasting JII.

Download Full-text

A Deep Learning based Arabic Script Recognition System: Benchmark on KHAT

The International Arab Journal of Information Technology ◽

10.34028/iajit/17/3/3 ◽

2020 ◽

Vol 17 (3) ◽

pp. 299-305 ◽

Cited By ~ 1

Author(s):

Riaz Ahmad ◽

Saeeda Naz ◽

Muhammad Afzal ◽

Sheikh Rashid ◽

Marcus Liwicki ◽

...

Keyword(s):

Deep Learning ◽

Character Recognition ◽

Data Augmentation ◽

Short Term Memory ◽

Recognition System ◽

Learning Approach ◽

Arabic Text ◽

Data Set ◽

Processing Step ◽

Handwritten Arabic

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.

Download Full-text