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

In this paper, we focus on the challenges of modeling deformable 3D objectsfrom casual videos. With the popularity of neural radiance fields (NeRF), manyworks extend it to dynamic scenes with a canonical NeRF and a deformation modelthat achieves 3D point transformation between the observation space and thecanonical space. Recent works rely on linear blend skinning (LBS) to achievethe canonical-observation transformation. However, the linearly weightedcombination of rigid transformation matrices is not guaranteed to be rigid. Asa matter of fact, unexpected scale and shear factors often appear. In practice,using LBS as the deformation model can always lead to skin-collapsing artifactsfor bending or twisting motions. To solve this problem, we propose neural dualquaternion blend skinning (NeuDBS) to achieve 3D point deformation, which canperform rigid transformation without skin-collapsing artifacts. In the endeavorto register 2D pixels across different frames, we establish a correspondencebetween canonical feature embeddings that encodes 3D points within thecanonical space, and 2D image features by solving an optimal transport problem.Besides, we introduce a texture filtering approach for texture rendering thateffectively minimizes the impact of noisy colors outside target deformableobjects. Extensive experiments on real and synthetic datasets show that ourapproach can reconstruct 3D models for humans and animals with betterqualitative and quantitative performance than state-of-the-art methods.

MoDA: Modeling Deformable 3D Objects from Casual Videos