Offline Handwritten Amharic Character Recognition Using Few-Shot Learning

Abstract

Few-shot learning is a demanding and challenging problem in machine learning aimed at learning from only fewer labeled training examples. It has become an active area of research due to deep learning requiring huge amounts of labeled dataset, which is not easy practically. Learning from fewer examples is also an important attempt towards learning like humans. Few-shot learning has proven a very good promise in different areas of machine learning applications, particularly in image classification. As it is a recent technique, most researchers focus on understanding and solving the issues related to its concept by focusing only on common image datasets like Mini-ImageNet and Omniglot. Few-shot learning also opens an opportunity to address low resource languages like Amharic. Hence in the current study, it is used to address offline handwritten character recognition for Amharic language. Particularly, prototypical networks, the popular and simpler type of few-shot learning, is implemented as a baseline. Using the opportunities explored in the nature of Amharic alphabet having columnwise and rowwise characters’ similarities, a novel way of augmenting the training episodes is proposed. According to the study’s experimental result, it was observed that the proposed method outperformed the baseline method in 1-shot setting. This study has implemented few-shot learning for Amharic characters for the first time. More importantly, the findings of the study open new ways of examining the influence of training episodes in few-shot learning, which is one of the critical issues requiring further exploration. The datasets used for this study are collected from native Amharic language writers using an Android App developed as a part of this study.