Text Recognition in the Wild: A Survey

TL;DR

This survey comprehensively reviews the evolution, current state, and future directions of scene text recognition, emphasizing deep learning advancements and providing valuable resources for researchers entering the field.

Contribution

It offers a complete overview of scene text recognition, including fundamental problems, recent innovations, resources, and future research directions.

Findings

Deep learning has significantly advanced scene text recognition.

The survey consolidates existing methods and resources.

Future research directions are identified for improving accuracy and efficiency.

Abstract

The history of text can be traced back over thousands of years. Rich and precise semantic information carried by text is important in a wide range of vision-based application scenarios. Therefore, text recognition in natural scenes has been an active research field in computer vision and pattern recognition. In recent years, with the rise and development of deep learning, numerous methods have shown promising in terms of innovation, practicality, and efficiency. This paper aims to (1) summarize the fundamental problems and the state-of-the-art associated with scene text recognition; (2) introduce new insights and ideas; (3) provide a comprehensive review of publicly available resources; (4) point out directions for future work. In summary, this literature review attempts to present the entire picture of the field of scene text recognition. It provides a comprehensive reference for people entering this field, and could be helpful to inspire future research. Related resources are available at our Github repository: https://github.com/HCIILAB/Scene-Text-Recognition.