computer-vision-1

Text Spotting

3260 papers • 126 benchmarks • 313 datasets

Text Spotting is the combination of Scene Text Detection and Scene Text Recognition in an end-to-end manner. It is the ability to read natural text in the wild.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in text-spotting-1

Trend

Dataset

Best Model

Actions

ICDAR 2015

Total-Text

SCUT-CTW1500

Libraries

i

Use these libraries to find text-spotting-1 models and implementations

hikopensource/davar-lab-ocr

4 papers 703

Datasets

Textual Visual Semantic Dataset

Subtasks

No subtasks available.

Most implemented papers

ABCNet: Real-Time Scene Text Spotting With Adaptive Bezier-Curve Network

Chunhua Shen, Tong He, Yuliang Liu, Lianwen Jin, Hao Chen, Liangwei Wang•Sun Feb 23 2020

For the first time, a novel BezierAlign layer is designed for extracting accurate convolution features of a text instance with arbitrary shapes, significantly improving the precision compared with previous methods and introducing negligible computation overhead.

385

Content

Inverse-Text

vitae-transformer/vitae-transformer…

2 papers 66

0

Paper Graph

FOTS: Fast Oriented Text Spotting with a Unified Network

Y. Qiao, Ding Liang, Junjie Yan, Xuebo Liu, Shipeng Yan, Dagui Chen•Thu Jan 04 2018

This work proposes a unified end-to-end trainable Fast Oriented Text Spotting (FOTS) network for simultaneous detection and recognition, sharing computation and visual information among the two complementary tasks, and introduces RoIRotate to share convolutional features between detection and Recognition.

524 0

Paper Graph

Visual Re-ranking with Natural Language Understanding for Text Spotting

F. Moreno-Noguer, Ahmed Sabir, Lluís Padró•Sun Oct 28 2018

This paper proposes a post-processing approach to improve scene text recognition accuracy by using occurrence probabilities of words (unigram language model), and the semantic correlation between scene and text.

15 0

Paper Graph

Semantic Relatedness Based Re-ranker for Text Spotting

F. Moreno-Noguer, Ahmed Sabir, Lluís Padró•Mon Sep 16 2019

It is shown how learning a word-to-word or word- to-sentence relatedness score can improve the performance of text spotting systems up to 2.9 points, outperforming other measures in a benchmark dataset.

5 0

Paper Graph

A Bilingual, OpenWorld Video Text Dataset and End-to-end Video Text Spotter with Transformer

Weijia Wu, Yuanqiang Cai, Debing Zhang, Sibo Wang, Zhuang Li, Jiahong Li, Yejun Tang, Hong Zhou•Wed Dec 08 2021

This work introduces a large-scale, Bilingual, Open World Video text benchmark dataset (BOVText), and proposes an end-to-end video text spotting framework with Transformer, termed TransVTSpotter, which solves the multi-orient text spotting in video with a simple, but efficient attention-based query-key mechanism.

38 0

Paper Graph

SPTS v2: Single-Point Scene Text Spotting

Chunhua Shen, Dahua Lin, Yuliang Liu, Lianwen Jin, Dezhi Peng, Xiang Bai, Jiaxin Zhang, Jingqun Tang, Mingxin Huang, Xinyu Wang, Can Huang•Tue Jan 03 2023

The SPTS v2 can outperform previous state-of-the-art single-point text spotters with fewer parameters while achieving 19× faster inference speed and experiments suggest a potential preference for single- point representation in scene text spotting when compared to other representations.

69 0

Paper Graph

ESTextSpotter: Towards Better Scene Text Spotting with Explicit Synergy in Transformer

Yuliang Liu, Lianwen Jin, Dezhi Peng, Xiang Bai, Jiaxin Zhang, Mingxin Huang, Can Huang, Hao Lu•Sat Aug 19 2023

This paper introduces a new model named Explicit Synergy-based Text Spotting Transformer framework (ESTextSpotter), which achieves explicit synergy by modeling discriminative and interactive features for text detection and recognition within a single decoder.

42 0

Paper Graph

A Feasible Framework for Arbitrary-Shaped Scene Text Recognition

Qingjie Liu, Yunhong Wang, Di Huang, Jinjin Zhang, Wei Wang•Mon Dec 09 2019

This paper proposes a feasible framework for multi-lingual arbitrary-shaped STR, including instance segmentation based text detection and language model based attention mechanism for text recognition.

4 0

Paper Graph

AE TextSpotter: Learning Visual and Linguistic Representation for Ambiguous Text Spotting

Chunhua Shen, Ding Liang, Wenhai Wang, Enze Xie, Tong Lu, P. Luo, Zhibo Yang, Xuebo Liu, Xiaozhong Ji•Sun Aug 02 2020

This work proposes a novel text spotter, named Ambiguity Eliminating Text Spotter (AE TextSpotter), which learns both visual and linguistic features to significantly reduce ambiguity in text detection, and is the first time to improve text detection by using a language model.

25 0

Paper Graph

PGNet: Real-time Arbitrarily-Shaped Text Spotting with Point Gathering Network

Errui Ding, Xiaoqiang Zhang, Pengfei Wang, Guangming Shi, Pengyuan Lyu, Junyu Han, Jingtuo Liu, Chengquan Zhang, Fei Qi, Shanshan Liu•Sun Apr 11 2021

The PGNet is a single-shot text spotter, where the pixel-level character classification map is learned with proposed PG-CTC loss avoiding the usage of character-level annotations, and a graph refinement module (GRM) is proposed to optimize the coarse recognition and improve the end-to-end performance.

95 0

Paper Graph

Adding a benchmark result helps the community track progress.