EdgeNet: A novel approach for Arabic numeral classification

TL;DR

This paper introduces EdgeNet, a deep learning model utilizing edge features for Arabic numeral classification, achieving 99.59% accuracy and addressing data diversity and learning effectiveness issues.

Contribution

The study merges and augments Arabic numeral datasets and proposes EdgeNet, a novel deep model leveraging edge features with residual connections for improved classification performance.

Findings

EdgeNet achieves 99.59% accuracy on validation data.

Unified dataset improves performance of existing models.

EdgeNet outperforms state-of-the-art methods.

Abstract

Despite the importance of handwritten numeral classification, a robust and effective method for a widely used language like Arabic is still due. This study focuses to overcome two major limitations of existing works: data diversity and effective learning method. Hence, the existing Arabic numeral datasets have been merged into a single dataset and augmented to introduce data diversity. Moreover, a novel deep model has been proposed to exploit diverse data samples of unified dataset. The proposed deep model utilizes the low-level edge features by propagating them through residual connection. To make a fair comparison with the proposed model, the existing works have been studied under the unified dataset. The comparison experiments illustrate that the unified dataset accelerates the performance of the existing works. Moreover, the proposed model outperforms the existing state-of-the-art Arabic handwritten numeral classification methods and obtain an accuracy of 99.59% in the validation phase. Apart from that, different state-of-the-art classification models have studied with the same dataset to reveal their feasibility for the Arabic numeral classification. Code available at http://github.com/sharif-apu/EdgeNet.