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

3D reconstruction of dynamic scenes is a long-standing problem in computergraphics and increasingly difficult the less information is available.Shape-from-Template (SfT) methods aim to reconstruct a template-based geometryfrom RGB images or video sequences, often leveraging just a single monocularcamera without depth information, such as regular smartphone recordings.Unfortunately, existing reconstruction methods are either unphysical and noisyor slow in optimization. To solve this problem, we propose a novel SfTreconstruction algorithm for cloth using a pre-trained neural surrogate modelthat is fast to evaluate, stable, and produces smooth reconstructions due to aregularizing physics simulation. Differentiable rendering of the simulated meshenables pixel-wise comparisons between the reconstruction and a target videosequence that can be used for a gradient-based optimization procedure toextract not only shape information but also physical parameters such asstretching, shearing, or bending stiffness of the cloth. This allows to retaina precise, stable, and smooth reconstructed geometry while reducing the runtimeby a factor of 400-500 compared to $\phi$-SfT, a state-of-the-art physics-basedSfT approach.

Physics-guided Shape-from-Template: Monocular Video Perception through Neural Surrogate Models