Abstract:Scene text detection remains a grand challenge due to the variation in text curvatures, orientations, and aspect ratios. One of the hardest problems in this task is how to represent text instances of arbitrary shapes. Although many methods have been proposed to model irregular texts in a flexible manner, most of them lose simplicity and robustness. Their complicated post-processings and the regression under Dirac delta distribution undermine the detection performance and the generalization ability. In this paper, we propose an efficient text instance representation named CentripetalText (CT), which decomposes text instances into the combination of text kernels and centripetal shifts. Specifically, we utilize the centripetal shifts to implement pixel aggregation, guiding the external text pixels to the internal text kernels. The relaxation operation is integrated into the dense regression for centripetal shifts, allowing the correct prediction in a range instead of a specific value. The convenient reconstruction of text contours and the tolerance of prediction errors in our method guarantee the high detection accuracy and the fast inference speed, respectively. Besides, we shrink our text detector into a proposal generation module, namely CentripetalText Proposal Network, replacing Segmentation Proposal Network in Mask TextSpotter v3 and producing more accurate proposals. To validate the effectiveness of our method, we conduct experiments on several commonly used scene text benchmarks, including both curved and multi-oriented text datasets. For the task of scene text detection, our approach achieves superior or competitive performance compared to other existing methods, e.g., F-measure of 86.3% at 40.0 FPS on Total-Text, F-measure of 86.1% at 34.8 FPS on MSRA-TD500, etc. For the task of end-to-end scene text recognition, our method outperforms Mask TextSpotter v3 by 1.1% on Total-Text.

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

TextField: Learning a Deep Direction Field for Irregular Scene Text Detection

Scene Text Detection with Fully Convolutional Neural Networks

CentripetalText: an Efficient Text Instance Representation for Scene Text Detection

CT-Net: Arbitrary-Shaped Text Detection via Contour Transformer

Irregular Scene Text Detection Via Attention Guided Border Labeling.

Multi-Orientation Scene Text Detection with Adaptive Clustering.

2D-CTC for Scene Text Recognition

Curved Text Detection in Natural Scene Images with Semi- and Weakly-Supervised Learning

Arbitrarily Shaped Scene Text Detection with a Mask Tightness Text Detector

ContourNet: Taking a Further Step Toward Accurate Arbitrary-shaped Scene Text Detection.

TextTubes for Detecting Curved Text in the Wild

Arbitrary Shape Scene Text Detection with Adaptive Text Region Representation

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

TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes

Curved Text Detection in Scenic Images via Proposal-Free Panoptic Segmentation and Deep Learning

Scene text detection and recognition: a survey

Tk-Text: Multi-Shaped Scene Text Detection Via Instance Segmentation

Follow the Curve: Arbitrarily Oriented Scene Text Detection Using Key Points Spotting and Curve Prediction

Character Region Awareness for Text Detection

Scene Text Detection via Holistic, Multi-Channel Prediction