Skeleton-Guided Artistic Text Recognition

doi:10.1007/978-3-032-04630-7_17

2026-01-01

PRE

DOI：10.1007/978-3-032-04630-7_17

原文链接

原文求助

摘要

En 中文

艺术文本识别因字体、视觉效果和背景噪声等美学设计元素的多样性而面临重大挑战，这些元素虽具美观性却损害识别精度。尽管字符的骨架结构在变化中保持稳定，现有方法未能充分利用此信息以实现鲁棒识别。本文提出骨架引导的艺术文本识别（SG-ATR）方法，通过显式利用骨架表示来增强文本特征提取。该框架整合骨架化模块以提取字符关键结构，同时抑制复杂背景的干扰。为推动艺术文本识别的基准测试，我们引入艺术文本野外数据集（ATTW），包含16,627个源自广泛认可来源、涵盖多元艺术风格的实例。大量实验表明，SG-ATR在ATTW上达到75.39%的准确率，超越现有最优（SOTA）方法。此外，SG-ATR在标准场景文本识别基准上表现良好，凸显其在非艺术文本场景下的鲁棒性。

Keyword:

Artistic Text Recognition

Scene Text Recognition

Skeleton-Guided Transformer

Artistic Text-In-The-Wild

期刊

DOCUMENT ANALYSIS AND RECOGNITION-ICDAR 2025, PT V

IF:

论文数:

被引数:

机构

vietnam national university ho chi minh city (vnuhcm) system

学者数:

7.2K

论文数: 4.2K

被引数: 8

引用论文

What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

IF0

2019-10-01

OAAI

Jeonghun Baek; Geewook Kim; Junyeop Lee; Sungrae Park; Dongyoon Han; Sangdoo Yun; Seong Joon Oh; Hwalsuk Lee

Total-Text: A Comprehensive Dataset for Scene Text Detection and Recognition

2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)

IF0

2017-11-01

OAAI

Chee Kheng Ch'ng; Chee Seng Chan

ICDAR 2015 competition on Robust Reading ICDAR 2015稳健阅读竞赛

2015 13th International Conference on Document Analysis and Recognition (ICDAR)

IF0

2015-08-01

PREAI

Dimosthenis Karatzas; Lluis Gomez-Bigorda; Anguelos Nicolaou; Suman Ghosh; Andrew Bagdanov; Masakazu Iwamura; Jiri Matas; Lukas Neumann; Vijay Ramaseshan Chandrasekhar; Shijian Lu; Faisal Shafait; Seiichi Uchida; Ernest Valveny

MORAN: A Multi-Object Rectified Attention Network for scene text recognition MORAN: 用于场景文本识别的多目标校正注意力网络

PATTERN RECOGNITION

IF7.6

2019-06-01

296

PREAI

Luo, Canjie; Jin, Lianwen; Sun, Zenghui

ICDAR 2013 Robust Reading Competition ICDAR 2013稳健阅读竞赛

2013 12th International Conference on Document Analysis and Recognition

IF0

2013-08-01

PREAI

Dimosthenis Karatzas; Faisal Shafait; Seiichi Uchida; Masakazu Iwamura; Lluis Gomez i Bigorda; Sergi Robles Mestre; Joan Mas; David Fernandez Mota; Jon Almazan Almazan; Lluis Pere de las Heras

Connectionist Temporal Classification

Studies in Computational Intelligence

IF0

2012-02-06

PREAI

Alex Graves

Mask TextSpotter: An End-to-End Trainable Neural Network for Spotting Text with Arbitrary Shapes

Lecture Notes in Computer Science

IF0

2018-10-09

PREAI

Pengyuan Lyu; Minghui Liao; Cong Yao; Wenhao Wu; Xiang Bai