A fine-grained approach to scene text script identification

TL;DR

This paper introduces a novel fine-grained method combining convolutional features and Naive-Bayes Nearest Neighbor for script identification in natural scene images, supported by a new benchmark dataset.

Contribution

It presents a new approach for script identification in scene text images and introduces a public dataset for joint text detection and script recognition.

Findings

Achieves state-of-the-art results on the new dataset

Generalizes well across different datasets and scripts

Supports multi-lingual scene text recognition in the wild

Abstract

This paper focuses on the problem of script identification in unconstrained scenarios. Script identification is an important prerequisite to recognition, and an indispensable condition for automatic text understanding systems designed for multi-language environments. Although widely studied for document images and handwritten documents, it remains an almost unexplored territory for scene text images. We detail a novel method for script identification in natural images that combines convolutional features and the Naive-Bayes Nearest Neighbor classifier. The proposed framework efficiently exploits the discriminative power of small stroke-parts, in a fine-grained classification framework. In addition, we propose a new public benchmark dataset for the evaluation of joint text detection and script identification in natural scenes. Experiments done in this new dataset demonstrate that the proposed method yields state of the art results, while it generalizes well to different datasets and variable number of scripts. The evidence provided shows that multi-lingual scene text recognition in the wild is a viable proposition. Source code of the proposed method is made available online.