Bengali Handwritten Grapheme Classification: Deep Learning Approach

TL;DR

This paper addresses Bengali handwritten grapheme classification by evaluating existing neural networks and proposing a new CNN model, achieving over 95% accuracy, and discusses future directions with RPN-VGGNet.

Contribution

The paper introduces a novel CNN model for Bengali grapheme classification and demonstrates its superior performance over existing models like MLP and ResNet50.

Findings

Proposed CNN achieved validation root accuracy of 95.32%.

Vowel and consonant accuracies exceeded 98%.

Explored RPN with VGGNet as a future enhancement.

Abstract

Despite being one of the most spoken languages in the world ($6^{th}$ based on population), research regarding Bengali handwritten grapheme (smallest functional unit of a writing system) classification has not been explored widely compared to other prominent languages. Moreover, the large number of combinations of graphemes in the Bengali language makes this classification task very challenging. With an effort to contribute to this research problem, we participate in a Kaggle competition \cite{kaggle_link} where the challenge is to separately classify three constituent elements of a Bengali grapheme in the image: grapheme root, vowel diacritics, and consonant diacritics. We explore the performances of some existing neural network models such as Multi-Layer Perceptron (MLP) and state of the art ResNet50. To further improve the performance we propose our own convolution neural network (CNN) model for Bengali grapheme classification with validation root accuracy 95.32\%, vowel accuracy 98.61\%, and consonant accuracy 98.76\%. We also explore Region Proposal Network (RPN) using VGGNet with a limited setting that can be a potential future direction to improve the performance.