Bengali Font Identification Using Deep Convolutional Neural Network

Abstract

Automatic font recognition or similar font suggestions from an image or picture are the core design works for many designers. This paper proposes a framework based on Convolution Neural Network (CNNs) to the widely neglected problem of bangla font recognition by the vision community. First of all, we build up the available large scale data set consisting of both labeled synthetic data by Adobe and partly labeled real-world data. Next to CNN is trained to classify images into predefined font classes. Global average pooling layer is proposed instead of fully connected layers over feature maps in the classification layer to correspondence between feature maps and output. Thus the feature maps can be easily interpreted as font categories confidence maps. We show that our method achieves state-of-the-art performance on a challenging data set of 10 selected bangla computer fonts with 96% line level accuracy. Large-scale experiments show that our approach is exceptionally viable on our synthetic test images and achieves promising results on real world test images.