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

Visual Question Answering (VQA) and Image Captioning (CAP), which are amongthe most popular vision-language tasks, have analogous scene-text versions thatrequire reasoning from the text in the image. Despite their obviousresemblance, the two are treated independently and, as we show, yieldtask-specific methods that can either see or read, but not both. In this work,we conduct an in-depth analysis of this phenomenon and propose UniTNT, aUnified Text-Non-Text approach, which grants existing multimodal architecturesscene-text understanding capabilities. Specifically, we treat scene-textinformation as an additional modality, fusing it with any pretrainedencoder-decoder-based architecture via designated modules. Thorough experimentsreveal that UniTNT leads to the first single model that successfully handlesboth task types. Moreover, we show that scene-text understanding capabilitiescan boost vision-language models' performance on general VQA and CAP by up to2.69% and 0.6 CIDEr, respectively.

Towards Models that Can See and Read