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

Our goal is to synthesize 3D human motions given textual inputs describingsimultaneous actions, for example 'waving hand' while 'walking' at the sametime. We refer to generating such simultaneous movements as performing 'spatialcompositions'. In contrast to temporal compositions that seek to transitionfrom one action to another, spatial compositing requires understanding whichbody parts are involved in which action, to be able to move themsimultaneously. Motivated by the observation that the correspondence betweenactions and body parts is encoded in powerful language models, we extract thisknowledge by prompting GPT-3 with text such as "what are the body partsinvolved in the action <action name>?", while also providing the parts list andfew-shot examples. Given this action-part mapping, we combine body parts fromtwo motions together and establish the first automated method to spatiallycompose two actions. However, training data with compositional actions isalways limited by the combinatorics. Hence, we further create synthetic datawith this approach, and use it to train a new state-of-the-art text-to-motiongeneration model, called SINC ("SImultaneous actioN Compositions for 3D humanmotions"). In our experiments, that training with such GPT-guided syntheticdata improves spatial composition generation over baselines. Our code ispublicly available at https://sinc.is.tue.mpg.de/.

SINC: Spatial Composition of 3D Human Motions for Simultaneous Action Generation