Original scientific paper

https://doi.org/10.17559/TV-20190319095323

A Novel Deep Convolutional Neural Network Architecture Based on Transfer Learning for Handwritten Urdu Character Recognition

Mohammed Aarif Kilvisharam Oziuddeen orcid.org/0000-0002-0848-6813 ; C. Abdul Hakeem College of Engineering & Technology, Melvisharam, Vellore, Tamil Nadu, India
Sivakumar Poruran ; Dr. N.G.P. Institute of Technology, Coimbatore, Tamil Nadu, India
Mohamed Yousuff Caffiyar ; C. Abdul Hakeem College of Engineering & Technology, Melvisharam, Vellore, Tamil Nadu, India

Abstract

Deep convolutional neural networks (CNN) have made a huge impact on computer vision and set the state-of-the-art in providing extremely definite classification results. For character recognition, where the training images are usually inadequate, mostly transfer learning of pre-trained CNN is often utilized. In this paper, we propose a novel deep convolutional neural network for handwritten Urdu character recognition by transfer learning three pre-trained CNN models. We fine-tuned the layers of these pre-trained CNNs so as to extract features considering both global and local details of the Urdu character structure. The extracted features from the three CNN models are concatenated to train with two fully connected layers for classification. The experiment is conducted on UNHD, EMILLE, DBAHCL, and CDB/Farsi dataset, and we achieve 97.18% average recognition accuracy which outperforms the individual CNNs and numerous conventional classification methods.

Keywords

deep learning; document analysis; pattern recognition; pretrained CNN; transfer learning; Urdu character recognition

Hrčak ID:

242316

URI

https://hrcak.srce.hr/242316

Publication date:

15.8.2020.

Visits: 2.168 *