Accurate Scene Text Detection through Border Semantics Awareness and Bootstrapping

TL;DR

This paper introduces a scene text detection method that combines bootstrapping and border semantics awareness to improve localization accuracy, especially for long words and text lines, achieving state-of-the-art results on multiple datasets.

Contribution

It proposes a novel bootstrapping approach for better training data utilization and a semantics-aware border detection technique for precise text localization.

Findings

Achieved 80.1 f-score on MSRA-TD500 dataset.

Outperformed existing methods on multiple benchmarks.

Enhanced localization accuracy for long text lines.

Abstract

This paper presents a scene text detection technique that exploits bootstrapping and text border semantics for accurate localization of texts in scenes. A novel bootstrapping technique is designed which samples multiple 'subsections' of a word or text line and accordingly relieves the constraint of limited training data effectively. At the same time, the repeated sampling of text 'subsections' improves the consistency of the predicted text feature maps which is critical in predicting a single complete instead of multiple broken boxes for long words or text lines. In addition, a semantics-aware text border detection technique is designed which produces four types of text border segments for each scene text. With semantics-aware text borders, scene texts can be localized more accurately by regressing text pixels around the ends of words or text lines instead of all text pixels which often leads to inaccurate localization while dealing with long words or text lines. Extensive experiments demonstrate the effectiveness of the proposed techniques, and superior performance is obtained over several public datasets, e. g. 80.1 f-score for the MSRA-TD500, 67.1 f-score for the ICDAR2017-RCTW, etc.