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

A key challenge of learning a visual representation for the 3D high fidelitygeometry of dressed humans lies in the limited availability of the ground truthdata (e.g., 3D scanned models), which results in the performance degradation of3D human reconstruction when applying to real-world imagery. We address thischallenge by leveraging a new data resource: a number of social media dancevideos that span diverse appearance, clothing styles, performances, andidentities. Each video depicts dynamic movements of the body and clothes of asingle person while lacking the 3D ground truth geometry. To learn a visualrepresentation from these videos, we present a new self-supervised learningmethod to use the local transformation that warps the predicted local geometryof the person from an image to that of another image at a different timeinstant. This allows self-supervision by enforcing a temporal coherence overthe predictions. In addition, we jointly learn the depths along with thesurface normals that are highly responsive to local texture, wrinkle, and shadeby maximizing their geometric consistency. Our method is end-to-end trainable,resulting in high fidelity depth estimation that predicts fine geometryfaithful to the input real image. We further provide a theoretical bound ofself-supervised learning via an uncertainty analysis that characterizes theperformance of the self-supervised learning without training. We demonstratethat our method outperforms the state-of-the-art human depth estimation andhuman shape recovery approaches on both real and rendered images.

Self-supervised 3D Representation Learning of Dressed Humans from Social Media Videos