https://arxiv.org/abs/1910.07954 In this paper, we have a convolutional character neural network where we have object detection by taking a character as a basic unit. First, we do character detection and recognition and then we go for text detection.
Here (Page number 5 under the subheading Iterative character detection) it is written that a model trained on English and Chinese texts will generalize well in regards to text detection. But how English and Chinese texts are good for generalization in text detection. If you have any queries regarding the paper you can ask me in the comment section Thanks in advance!