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

Recently, 3D vision-and-language tasks have attracted increasing researchinterest. Compared to other vision-and-language tasks, the 3D visual questionanswering (VQA) task is less exploited and is more susceptible to languagepriors and co-reference ambiguity. Meanwhile, a couple of recently proposed 3DVQA datasets do not well support 3D VQA task due to their limited scale andannotation methods. In this work, we formally define and address a 3D groundedVQA task by collecting a new 3D VQA dataset, referred to as FE-3DGQA, withdiverse and relatively free-form question-answer pairs, as well as dense andcompletely grounded bounding box annotations. To achieve more explainableanswers, we labelled the objects appeared in the complex QA pairs withdifferent semantic types, including answer-grounded objects (both appeared andnot appeared in the questions), and contextual objects for answer-groundedobjects. We also propose a new 3D VQA framework to effectively predict thecompletely visually grounded and explainable answer. Extensive experimentsverify that our newly collected benchmark datasets can be effectively used toevaluate various 3D VQA methods from different aspects and our newly proposedframework also achieves state-of-the-art performance on the new benchmarkdataset. Both the newly collected dataset and our codes will be publiclyavailable at http://github.com/zl/3DGQA.

Towards Explainable 3D Grounded Visual Question Answering: A New Benchmark and Strong Baseline