Multi-script Handwritten Digit Recognition Using Multi-task Learning

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.

Improving Amharic Handwritten Word Recognition Using Auxiliary Task

“…Finally, relevant input to this study is explored in the work of Gondere et al [10] on how auxiliary tasks are extracted from the Amharic alphabet and improve Amharic character recognition using muli-task learning. A related and typical example of how such a way of addressing specific problem help in innovative and generalized problem solution for multi-script handwritten digit recognition is demonstrated in [11].…”

“…It should be noted that paths should be mapped using a function (F ) into correct label sequences by removing the repeated symbols and blanks for providing the correct label path for the set of aligned paths. Further details about the baseline method can be found in [1] and a broader insight of the proposed model is presented in [10,11].…”