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

End-to-end scene text spotting, which aims to read the text in naturalimages, has garnered significant attention in recent years. However, recentstate-of-the-art methods usually incorporate detection and recognition simplyby sharing the backbone, which does not directly take advantage of the featureinteraction between the two tasks. In this paper, we propose a new end-to-endscene text spotting framework termed SwinTextSpotter v2, which seeks to find abetter synergy between text detection and recognition. Specifically, we enhancethe relationship between two tasks using novel Recognition Conversion andRecognition Alignment modules. Recognition Conversion explicitly guides textlocalization through recognition loss, while Recognition Alignment dynamicallyextracts text features for recognition through the detection predictions. Thissimple yet effective design results in a concise framework that requiresneither an additional rectification module nor character-level annotations forthe arbitrarily-shaped text. Furthermore, the parameters of the detector aregreatly reduced without performance degradation by introducing a Box SelectionSchedule. Qualitative and quantitative experiments demonstrate thatSwinTextSpotter v2 achieved state-of-the-art performance on variousmultilingual (English, Chinese, and Vietnamese) benchmarks. The code will beavailable at\href{https://github.com/mxin262/SwinTextSpotterv2}{SwinTextSpotter v2}.

SwinTextSpotter v2: Towards Better Synergy for Scene Text Spotting