Handwritten Gurmukhi Digit Recognition System for Small Datasets

Received: 6 May 2020 Accepted: 10 August 2020 In the present study, a method to increase the recognition accuracy of Gurmukhi (Indian Regional Script) Handwritten Digits has been proposed. The proposed methodology uses a DCNN (Deep Convolutional Neural Network) with a cascaded XGBoost (Extreme Gradient Boosting) algorithm. Also, a comprehensive analysis has been done to apprehend the impact of kernel size of DCNN on recognition accuracy. The reason for using DCNN is its impressive performance in terms of recognition accuracy of handwritten digits, but in order to achieve good recognition accuracy, DCNN requires a huge amount of data and also significant training/testing time. In order to increase the accuracy of DCNN for a small dataset more images have been generated by applying a shear transformation (A transformation that preserves parallelism but not length and angles) to the original images. To address the issue of large training time only two hidden layers along with selective cascading XGBoost among the misclassified digits have been used. Also, the issue of overfitting is discussed in detail and has been reduced to a great extent. Finally, the results are compared with performance of some recent techniques like SVM (Support Vector Machine) Random Forest, and XGBoost classifiers on DCT (Discrete Cosine Transform) and DWT (Discrete Wavelet Transform) features obtained on the same dataset. It is found that proposed methodology can outperform other techniques in terms of overall rate of recognition.

[1]  Lawrence D. Jackel,et al.  Backpropagation Applied to Handwritten Zip Code Recognition , 1989, Neural Computation.

[2]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[3]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[4]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[5]  Yoshua. Bengio,et al.  Learning Deep Architectures for AI , 2007, Found. Trends Mach. Learn..

[6]  Luca Maria Gambardella,et al.  Deep, Big, Simple Neural Nets for Handwritten Digit Recognition , 2010, Neural Computation.

[7]  Singh Siddharth Kartar,et al.  Handwritten Gurmukhi Numeral Recognition using Different Feature Sets , 2011 .

[8]  Munish Kumar,et al.  Classification of characters and grading writers in offline handwritten Gurmukhi script , 2011, 2011 International Conference on Image Information Processing.

[9]  Rajendra Kumar Sharma,et al.  A Novel Feature Extraction Technique for Offline Handwritten Gurmukhi Character Recognition , 2013 .

[10]  Gurpreet Singh,et al.  Feature Extraction of Gurmukhi Script and Numerals: A Review of Offline Techniques , 2013 .

[11]  Avneet Kaur,et al.  Punjabi Handwritten Character Recognition Using Wavelet Based Features , 2015 .

[12]  Chokri Ben Amar,et al.  Dyadic Multi-resolution Analysis-Based Deep Learning for Arabic Handwritten Character Classification , 2015, 2015 IEEE 27th International Conference on Tools with Artificial Intelligence (ICTAI).

[13]  Shailesh Acharya,et al.  Deep learning based large scale handwritten Devanagari character recognition , 2015, 2015 9th International Conference on Software, Knowledge, Information Management and Applications (SKIMA).

[14]  Gurpreet Singh,et al.  Offline Gurmukhi script recognition using knowledge based approach & Multi-Layered Perceptron neural network , 2015, 2015 International Conference on Signal Processing, Computing and Control (ISPCC).

[15]  Keiron O'Shea.  Massively Deep Artificial Neural Networks for Handwritten Digit Recognition , 2015, ArXiv.

[16]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[17]  Monji Kherallah,et al.  Feature Extractor Based Deep Method to Enhance Online Arabic Handwritten Recognition System , 2016, ICANN.

[18]  Hang Li,et al.  Very Deep Neural Network for Handwritten Digit Recognition , 2016, IDEAL.

[19]  B. Chandra,et al.  Fast learning in Deep Neural Networks , 2016, Neurocomputing.

[20]  Xinyi Yang,et al.  Deep Convolutional Extreme Learning Machine and Its Application in Handwritten Digit Classification , 2016, Comput. Intell. Neurosci..

[21]  Narendra S. Chaudhari,et al.  Deep Convolutional Neural Network Classifier for Handwritten Devanagari Character Recognition , 2016 .

[22]  Mrutyunjaya Panda,et al.  Towards the effectiveness of Deep Convolutional Neural Network based Fast Random Forest Classifier , 2016, ArXiv.

[23]  Mohammed Kamel Benkaddour,et al.  Feature extraction and classification using deep convolutional neural networks, PCA and SVC for face recognition , 2017, Traitement du Signal.

[24]  Rajendra Kumar Sharma,et al.  Offline Handwritten Gurmukhi Character Recognition: Analytical Study of Different Transformations , 2017 .

[25]  A. Pravin,et al.  Ensembling: Model of histogram of oriented gradient based handwritten devanagari character recognition system , 2017, Traitement du Signal.

[26]  Ritika Wason,et al.  Deep learning: Evolution and expansion , 2018, Cognitive Systems Research.

[27]  Ramesh Neelapu,et al.  Deep learning based conventional neural network architecture for medical image classification , 2018, Traitement du Signal.

[28]  Jaafar Alghazo,et al.  Multi-Language Handwritten Digits Recognition based on Novel Structural Features , 2019, Journal of Imaging Science and Technology.

[29]  N. Veeranjaneyulu,et al.  Feature Extraction and Classification Using Deep Convolutional Neural Networks , 2019 .

[30]  Huihuang Zhao,et al.  Multiple classifiers fusion and CNN feature extraction for handwritten digits recognition , 2020, Granular Computing.

[31]  Kaizhu Huang,et al.  Offline Arabic Handwriting Recognition Using Deep Machine Learning: A Review of Recent Advances , 2019, BICS.

[32]  Asif Ali Wagan,et al.  Handwritten Urdu character recognition via images using different machine learning and deep learning techniques , 2020 .

[33]  Semiye Demircan,et al.  Comparison of the Effects of Mel Coefficients and Spectrogram Images via Deep Learning in Emotion Classification , 2020, Traitement du Signal.

[34]  R. Ramya,et al.  Design and simulation of handwritten recognition system , 2020 .