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

Vision model have gained increasing attention due to their simplicity andefficiency in Scene Text Recognition (STR) task. However, due to lacking theperception of linguistic knowledge and information, recent vision models sufferfrom two problems: (1) the pure vision-based query results in attention drift,which usually causes poor recognition and is summarized as linguisticinsensitive drift (LID) problem in this paper. (2) the visual feature issuboptimal for the recognition in some vision-missing cases (e.g. occlusion,etc.). To address these issues, we propose a $\textbf{L}$inguistic$\textbf{P}$erception $\textbf{V}$ision model (LPV), which explores thelinguistic capability of vision model for accurate text recognition. Toalleviate the LID problem, we introduce a Cascade Position Attention (CPA)mechanism that obtains high-quality and accurate attention maps throughstep-wise optimization and linguistic information mining. Furthermore, a GlobalLinguistic Reconstruction Module (GLRM) is proposed to improve therepresentation of visual features by perceiving the linguistic information inthe visual space, which gradually converts visual features into semanticallyrich ones during the cascade process. Different from previous methods, ourmethod obtains SOTA results while keeping low complexity (92.4% accuracy withonly 8.11M parameters). Code is available athttps://github.com/CyrilSterling/LPV.

Linguistic More: Taking a Further Step toward Efficient and Accurate Scene Text Recognition