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

In recent years, end-to-end scene text spotting approaches are evolving tothe Transformer-based framework. While previous studies have shown the crucialimportance of the intrinsic synergy between text detection and recognition,recent advances in Transformer-based methods usually adopt an implicit synergystrategy with shared query, which can not fully realize the potential of thesetwo interactive tasks. In this paper, we argue that the explicit synergyconsidering distinct characteristics of text detection and recognition cansignificantly improve the performance text spotting. To this end, we introducea new model named Explicit Synergy-based Text Spotting Transformer framework(ESTextSpotter), which achieves explicit synergy by modeling discriminative andinteractive features for text detection and recognition within a singledecoder. Specifically, we decompose the conventional shared query intotask-aware queries for text polygon and content, respectively. Through thedecoder with the proposed vision-language communication module, the queriesinteract with each other in an explicit manner while preserving discriminativepatterns of text detection and recognition, thus improving performancesignificantly. Additionally, we propose a task-aware query initializationscheme to ensure stable training. Experimental results demonstrate that ourmodel significantly outperforms previous state-of-the-art methods. Code isavailable at https://github.com/mxin262/ESTextSpotter.

ESTextSpotter: Towards Better Scene Text Spotting with Explicit Synergy in Transformer