Multi-script Handwritten Digit Recognition Using Multi-task Learning

TL;DR

This paper explores multi-script handwritten digit recognition using multi-task learning, demonstrating a novel approach that improves accuracy and simplifies loss balancing across Latin, Arabic, Kannada, and Amharic scripts.

Contribution

It introduces a new method for multi-task learning that enhances classification performance without requiring loss weighting, applied to multiple scripts including Amharic.

Findings

Proposed method outperforms baseline models.

Avoids need for loss weighting in multi-task learning.

Shows promising results on Latin, Arabic, Kannada, and Amharic scripts.

Abstract

Handwritten digit recognition is one of the extensively studied area in machine learning. Apart from the wider research on handwritten digit recognition on MNIST dataset, there are many other research works on various script recognition. However, it is not very common for multi-script digit recognition which encourage the development of robust and multipurpose systems. Additionally working on multi-script digit recognition enables multi-task learning, considering the script classification as a related task for instance. It is evident that multi-task learning improves model performance through inductive transfer using the information contained in related tasks. Therefore, in this study multi-script handwritten digit recognition using multi-task learning will be investigated. As a specific case of demonstrating the solution to the problem, Amharic handwritten character recognition will also be experimented. The handwritten digits of three scripts including Latin, Arabic and Kannada are studied to show that multi-task models with reformulation of the individual tasks have shown promising results. In this study a novel way of using the individual tasks predictions was proposed to help classification performance and regularize the different loss for the purpose of the main task. This finding has outperformed the baseline and the conventional multi-task learning models. More importantly, it avoided the need for weighting the different losses of the tasks, which is one of the challenges in multi-task learning.