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

Generating coherent and useful image/video scenes from a free-form textualdescription is technically a very difficult problem to handle. Textualdescription of the same scene can vary greatly from person to person, orsometimes even for the same person from time to time. As the choice of wordsand syntax vary while preparing a textual description, it is challenging forthe system to reliably produce a consistently desirable output from differentforms of language input. The prior works of scene generation have been mostlyconfined to rigorous sentence structures of text input which restrict thefreedom of users to write description. In our work, we study a new pipelinethat aims to generate static as well as animated 3D scenes from different typesof free-form textual scene description without any major restriction. Inparticular, to keep our study practical and tractable, we focus on a smallsubspace of all possible 3D scenes, containing various combinations of cube,cylinder and sphere. We design a two-stage pipeline. In the first stage, weencode the free-form text using an encoder-decoder neural architecture. In thesecond stage, we generate a 3D scene based on the generated encoding. Ourneural architecture exploits state-of-the-art language model as encoder toleverage rich contextual encoding and a new multi-head decoder to predictmultiple features of an object in the scene simultaneously. For ourexperiments, we generate a large synthetic data-set which contains 13,00,000and 14,00,000 samples of unique static and animated scene descriptions,respectively. We achieve 98.427% accuracy on test data set in detecting the 3Dobjects features successfully. Our work shows a proof of concept of oneapproach towards solving the problem, and we believe with enough training data,the same pipeline can be expanded to handle even broader set of 3D scenegeneration problems.

Static and Animated 3D Scene Generation from Free-form Text Descriptions