Image Pre-processing on NumtaDB for Bengali Handwritten Digit Recognition
This work addresses a domain-specific problem for researchers and practitioners in Bengali digit recognition by providing a benchmark, though it is incremental as it focuses on pre-processing rather than novel model development.
The paper tackled the lack of pre-processed data for Bengali handwritten digit recognition by establishing a benchmark for image pre-processing on the NumtaDB dataset, aiming to improve accuracy across machine learning models.
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.