… In this paper, we revisit scene text recognition from a data perspective. Despite the current
benchmarks being close to saturation, we argue that the problem of STR remains unsolved, …

… of text into a one-dimensional form may lose useful information and introduce extra noise.
In this paper, we approach scene text recognition from a two-dimensional perspective. A …

… data augmentation is crucial to scene text recognition … scene text recognition as an image
classification problem. Based on the image classification perspective, we design a scene text …

… manner of describing and modelling text and background in natural scenes. In this paper, …
on scene text detection and recognition in the past few years, mainly from the perspective of …

… and data augmentation. In this paper, our focus is on the relatively straightforward
approach of data augmentation. In STR, data augmentation has not been rigorously explored. …

… as script identification, text/non-text classification and text-to-… directions on scene text detection
and recognition that need … [40]; in order to recognize perspective scene texts of arbitrary …

… -scale scene text dataset of Chinese shop signs in street views. It contains 25,770 natural …
on scene text detection. Moreover, by comparing the accuracy of four scene text recognition …

… We review methods from different perspectives and list the up-to-date datasets. We also …
that scene text detection can be taxonomically subsumed under general object detection, which …

… We show that the latter approach achieves superior performance. While scene text recognition
has generally been … Adopting this approach opens the door for real world scene text …

… an approach to text recognition in natural scene images. … present an approach to recognize
perspective scene texts of … reduces the cost of collecting training data. (3) For performance …