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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent methods for dynamic human reconstruction have attained promisingreconstruction results. Most of these methods rely only on RGB colorsupervision without considering explicit geometric constraints. This leads toexisting human reconstruction techniques being more prone to overfitting tocolor and causes geometrically inherent ambiguities, especially in the sparsemulti-view setup.  Motivated by recent advances in the field of monocular geometry prediction,we consider the geometric constraints of estimated depth and normals in thelearning of neural implicit representation for dynamic human reconstruction. Asa geometric regularization, this provides reliable yet explicit supervisioninformation, and improves reconstruction quality. We also exploit severalbeneficial physical priors, such as adding noise into view direction andmaximizing the density on the human surface. These priors ensure the colorrendered along rays to be robust to view direction and reduce the inherentambiguities of density estimated along rays. Experimental results demonstratethat depth and normal cues, predicted by human-specific monocular estimators,can provide effective supervision signals and render more accurate images.Finally, we also show that the proposed physical priors significantly reduceoverfitting and improve the overall quality of novel view synthesis. Our codeis availableat:~\href{https://github.com/PRIS-CV/HumanRecon}{https://github.com/PRIS-CV/HumanRecon}.

HumanRecon: Neural Reconstruction of Dynamic Human Using Geometric Cues and Physical Priors