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

Inspired by the great success of language model (LM)-based pre-training,recent studies in visual document understanding have explored LM-basedpre-training methods for modeling text within document images. Among them,pre-training that reads all text from an image has shown promise, but oftenexhibits instability and even fails when applied to broader domains, such asthose involving both visual documents and scene text images. This is asubstantial limitation for real-world scenarios, where the processing of textimage inputs in diverse domains is essential. In this paper, we investigateeffective pre-training tasks in the broader domains and also propose a novelpre-training method called SCOB that leverages character-wise supervisedcontrastive learning with online text rendering to effectively pre-traindocument and scene text domains by bridging the domain gap. Moreover, SCOBenables weakly supervised learning, significantly reducing annotation costs.Extensive benchmarks demonstrate that SCOB generally improves vanillapre-training methods and achieves comparable performance to state-of-the-artmethods. Our findings suggest that SCOB can be served generally and effectivelyfor read-type pre-training methods. The code will be available athttps://github.com/naver-ai/scob.

SCOB: Universal Text Understanding via Character-wise Supervised Contrastive Learning with Online Text Rendering for Bridging Domain Gap