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

Text-VQA aims at answering questions that require understanding the textualcues in an image. Despite the great progress of existing Text-VQA methods,their performance suffers from insufficient human-labeled question-answer (QA)pairs. However, we observe that, in general, the scene text is not fullyexploited in the existing datasets -- only a small portion of the text in eachimage participates in the annotated QA activities. This results in a huge wasteof useful information. To address this deficiency, we develop a new method togenerate high-quality and diverse QA pairs by explicitly utilizing the existingrich text available in the scene context of each image. Specifically, wepropose, TAG, a text-aware visual question-answer generation architecture thatlearns to produce meaningful, and accurate QA samples using a multimodaltransformer. The architecture exploits underexplored scene text information andenhances scene understanding of Text-VQA models by combining the generated QApairs with the initial training data. Extensive experimental results on twowell-known Text-VQA benchmarks (TextVQA and ST-VQA) demonstrate that ourproposed TAG effectively enlarges the training data that helps improve theText-VQA performance without extra labeling effort. Moreover, our modeloutperforms state-of-the-art approaches that are pre-trained with extralarge-scale data. Code is available at https://github.com/HenryJunW/TAG.

TAG: Boosting Text-VQA via Text-aware Visual Question-answer Generation