Handwritten Isolated Bangla Compound Character Recognition: a new benchmark using a novel deep learning approach

TL;DR

This paper introduces a novel deep learning approach using layerwise training of DCNNs for recognizing handwritten Bangla compound characters, setting a new accuracy benchmark on the CMATERdb dataset.

Contribution

It presents a new deep learning technique with supervised layerwise training and RMSProp optimization that outperforms existing shallow and standard deep models.

Findings

Achieved a recognition error rate of 9.67%.

Set a new benchmark accuracy of 90.33%.

Outperformed traditional shallow learning methods.

Abstract

In this work, a novel deep learning technique for the recognition of handwritten Bangla isolated compound character is presented and a new benchmark of recognition accuracy on the CMATERdb 3.1.3.3 dataset is reported. Greedy layer wise training of Deep Neural Network has helped to make significant strides in various pattern recognition problems. We employ layerwise training to Deep Convolutional Neural Networks (DCNN) in a supervised fashion and augment the training process with the RMSProp algorithm to achieve faster convergence. We compare results with those obtained from standard shallow learning methods with predefined features, as well as standard DCNNs. Supervised layerwise trained DCNNs are found to outperform standard shallow learning models such as Support Vector Machines as well as regular DCNNs of similar architecture by achieving error rate of 9.67% thereby setting a new benchmark on the CMATERdb 3.1.3.3 with recognition accuracy of 90.33%, representing an improvement of nearly 10%.