Image Pre-processing on NumtaDB for Bengali Handwritten Digit Recognition

TL;DR

This paper aims to establish effective image pre-processing benchmarks for Bengali handwritten digit recognition using NumtaDB, addressing the lack of pre-processed data for improved machine learning model accuracy.

Contribution

It introduces pre-processing benchmarks for Bengali digits in NumtaDB, facilitating better model performance and providing a foundation for future research in Bengali handwritten digit recognition.

Findings

Identified effective pre-processing techniques for Bengali digits

Established benchmark accuracy levels for various models

Enhanced recognition performance on the NumtaDB dataset

Abstract

NumtaDB is by far the largest data-set collection for handwritten digits in Bengali. This is a diverse dataset containing more than 85000 images. But this diversity also makes this dataset very difficult to work with. The goal of this paper is to find the benchmark for pre-processed images which gives good accuracy on any machine learning models. The reason being, there are no available pre-processed data for Bengali digit recognition to work with like the English digits for MNIST.