Inter- and Intra-domain Knowledge Transfer for Related Tasks in Deep Character Recognition

TL;DR

This paper investigates the effectiveness of deep transfer learning for character recognition, finding that transfer learning does not always outperform traditional methods, especially across varying task similarities.

Contribution

The study provides a comprehensive analysis of inter- and intra-domain transfer learning effects on character recognition, highlighting scenarios where transfer learning offers limited benefits.

Findings

Transfer learning does not significantly outperform traditional methods in character recognition.

The effectiveness of transfer learning depends on the similarity between source and target tasks.

Parameter and feature transfer behaviors vary with task similarity.

Abstract

Pre-training a deep neural network on the ImageNet dataset is a common practice for training deep learning models, and generally yields improved performance and faster training times. The technique of pre-training on one task and then retraining on a new one is called transfer learning. In this paper we analyse the effectiveness of using deep transfer learning for character recognition tasks. We perform three sets of experiments with varying levels of similarity between source and target tasks to investigate the behaviour of different types of knowledge transfer. We transfer both parameters and features and analyse their behaviour. Our results demonstrate that no significant advantage is gained by using a transfer learning approach over a traditional machine learning approach for our character recognition tasks. This suggests that using transfer learning does not necessarily presuppose a better performing model in all cases.