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

Learning based representation has become the key to the success of manycomputer vision systems. While many 3D representations have been proposed, itis still an unaddressed problem how to represent a dynamically changing 3Dobject. In this paper, we introduce a compositional representation for 4Dcaptures, i.e. a deforming 3D object over a temporal span, that disentanglesshape, initial state, and motion respectively. Each component is represented bya latent code via a trained encoder. To model the motion, a neural OrdinaryDifferential Equation (ODE) is trained to update the initial state conditionedon the learned motion code, and a decoder takes the shape code and the updatedstate code to reconstruct the 3D model at each time stamp. To this end, wepropose an Identity Exchange Training (IET) strategy to encourage the networkto learn effectively decoupling each component. Extensive experimentsdemonstrate that the proposed method outperforms existing state-of-the-art deeplearning based methods on 4D reconstruction, and significantly improves onvarious tasks, including motion transfer and completion.

Learning Compositional Representation for 4D Captures with Neural ODE