English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Scene text detection remains a grand challenge due to the variation in textcurvatures, orientations, and aspect ratios. One of the hardest problems inthis task is how to represent text instances of arbitrary shapes. Although manymethods have been proposed to model irregular texts in a flexible manner, mostof them lose simplicity and robustness. Their complicated post-processings andthe regression under Dirac delta distribution undermine the detectionperformance and the generalization ability. In this paper, we propose anefficient text instance representation named CentripetalText (CT), whichdecomposes text instances into the combination of text kernels and centripetalshifts. Specifically, we utilize the centripetal shifts to implement pixelaggregation, guiding the external text pixels to the internal text kernels. Therelaxation operation is integrated into the dense regression for centripetalshifts, allowing the correct prediction in a range instead of a specific value.The convenient reconstruction of text contours and the tolerance of predictionerrors in our method guarantee the high detection accuracy and the fastinference speed, respectively. Besides, we shrink our text detector into aproposal generation module, namely CentripetalText Proposal Network, replacingSegmentation Proposal Network in Mask TextSpotter v3 and producing moreaccurate proposals. To validate the effectiveness of our method, we conductexperiments on several commonly used scene text benchmarks, including bothcurved and multi-oriented text datasets. For the task of scene text detection,our approach achieves superior or competitive performance compared to otherexisting methods, e.g., F-measure of 86.3% at 40.0 FPS on Total-Text, F-measureof 86.1% at 34.8 FPS on MSRA-TD500, etc. For the task of end-to-end scene textrecognition, our method outperforms Mask TextSpotter v3 by 1.1% on Total-Text.

CentripetalText: An Efficient Text Instance Representation for Scene Text Detection