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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Document dewarping, aiming to eliminate geometric deformation in photographeddocuments to benefit text recognition, has made great progress in recent yearsbut is still far from being solved. While Cartesian coordinates are typicallyleveraged by state-of-the-art approaches to learn a group of deformationcontrol points, such representation is not efficient for dewarping model tolearn the deformation information. In this work, we explore Polar coordinatesrepresentation for each point in document dewarping, namely Polar-Doc. Incontrast to most current works adopting a two-stage pipeline typically, Polarrepresentation enables a unified point regression framework for bothsegmentation and dewarping network in one single stage. Such unification makesthe whole model more efficient to learn under an end-to-end optimizationpipeline, and also obtains a compact representation. Furthermore, we propose anovel multi-scope Polar-Doc-IOU loss to constrain the relationship amongcontrol points as a grid-based regularization under the Polar representation.Visual comparisons and quantitative experiments on two benchmarks show that,with much fewer parameters than the other mainstream counterparts, ourone-stage model with multi-scope constraints achieves new state-of-the-artperformance on both pixel alignment metrics and OCR metrics. Source codes willbe available at \url{}.

Polar-Doc: One-Stage Document Dewarping with Multi-Scope Constraints under Polar Representation