# Accurate Scene Text Detection through Border Semantics Awareness and Bootstrapping

arXiv: https://arxiv.org/abs/1807.03547 · HTML: https://mlsanity.com/paper/1807.03547/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jul 2018.

## Built on by (most cited first, 20 of 20)

- [Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion](https://mlsanity.com/paper/2202.10304/) (2022)
- [Unbalanced Feature Transport for Exemplar-based Image Translation](https://mlsanity.com/paper/2106.10482/) (2021)
- [GMLight: Lighting Estimation via Geometric Distribution Approximation](https://mlsanity.com/paper/2102.10244/) (2021)
- [Arbitrary Shape Text Detection via Boundary Transformer](https://mlsanity.com/paper/2205.05320/) (2022)
- [Language Matters: A Weakly Supervised Vision-Language Pre-training Approach for Scene Text Detection and Spotting](https://mlsanity.com/paper/2203.03911/) (2022)
- [TPSNet: Reverse Thinking of Thin Plate Splines for Arbitrary Shape Scene Text Representation](https://mlsanity.com/paper/2110.12826/) (2021)
- [Mask is All You Need: Rethinking Mask R-CNN for Dense and Arbitrary-Shaped Scene Text Detection](https://mlsanity.com/paper/2109.03426/) (2021)
- [Sparse Needlets for Lighting Estimation with Spherical Transport Loss](https://mlsanity.com/paper/2106.13090/) (2021)
- [Arbitrary Shape Text Detection via Segmentation with Probability Maps](https://mlsanity.com/paper/2208.12419/) (2022)
- [Towards Robust Real-Time Scene Text Detection: From Semantic to Instance Representation Learning](https://mlsanity.com/paper/2308.07202/) (2023)
- [Graph Fusion Network for Multi-Oriented Object Detection](https://mlsanity.com/paper/2205.03562/) (2022)
- [Contextual Text Block Detection towards Scene Text Understanding](https://mlsanity.com/paper/2207.12955/) (2022)
- [Detection and Rectification of Arbitrary Shaped Scene Texts by using Text Keypoints and Links](https://mlsanity.com/paper/2103.00785/) (2021)
- [CORE-Text: Improving Scene Text Detection with Contrastive Relational Reasoning](https://mlsanity.com/paper/2112.07513/) (2021)
- [A Survey on Deep Domain Adaptation and Tiny Object Detection Challenges, Techniques and Datasets](https://mlsanity.com/paper/2107.07927/) (2021)
- [Domain Adaptive Scene Text Detection via Subcategorization](https://mlsanity.com/paper/2212.00377/) (2022)
- [Scene Text recognition with Full Normalization](https://mlsanity.com/paper/2109.01034/) (2021)
- [DPTNet: A Dual-Path Transformer Architecture for Scene Text Detection](https://mlsanity.com/paper/2208.09878/) (2022)
- [Research on Multilingual Natural Scene Text Detection Algorithm](https://mlsanity.com/paper/2312.11153/) (2023)
- [Invizo: Arabic Handwritten Document Optical Character Recognition Solution](https://mlsanity.com/paper/2502.05277/) (2025)

## Abstract (arXiv)

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.