RigGS: Rigging of 3D Gaussians for Modeling Articulated Objects in Videos

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yao, Yuxin, Deng, Zhi, Hou, Junhui
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913749666889728
author Yao, Yuxin
Deng, Zhi
Hou, Junhui
author_facet Yao, Yuxin
Deng, Zhi
Hou, Junhui
contents This paper considers the problem of modeling articulated objects captured in 2D videos to enable novel view synthesis, while also being easily editable, drivable, and re-posable. To tackle this challenging problem, we propose RigGS, a new paradigm that leverages 3D Gaussian representation and skeleton-based motion representation to model dynamic objects without utilizing additional template priors. Specifically, we first propose skeleton-aware node-controlled deformation, which deforms a canonical 3D Gaussian representation over time to initialize the modeling process, producing candidate skeleton nodes that are further simplified into a sparse 3D skeleton according to their motion and semantic information. Subsequently, based on the resulting skeleton, we design learnable skin deformations and pose-dependent detailed deformations, thereby easily deforming the 3D Gaussian representation to generate new actions and render further high-quality images from novel views. Extensive experiments demonstrate that our method can generate realistic new actions easily for objects and achieve high-quality rendering.
format Preprint
id arxiv_https___arxiv_org_abs_2503_16822
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RigGS: Rigging of 3D Gaussians for Modeling Articulated Objects in Videos
Yao, Yuxin
Deng, Zhi
Hou, Junhui
Computer Vision and Pattern Recognition
This paper considers the problem of modeling articulated objects captured in 2D videos to enable novel view synthesis, while also being easily editable, drivable, and re-posable. To tackle this challenging problem, we propose RigGS, a new paradigm that leverages 3D Gaussian representation and skeleton-based motion representation to model dynamic objects without utilizing additional template priors. Specifically, we first propose skeleton-aware node-controlled deformation, which deforms a canonical 3D Gaussian representation over time to initialize the modeling process, producing candidate skeleton nodes that are further simplified into a sparse 3D skeleton according to their motion and semantic information. Subsequently, based on the resulting skeleton, we design learnable skin deformations and pose-dependent detailed deformations, thereby easily deforming the 3D Gaussian representation to generate new actions and render further high-quality images from novel views. Extensive experiments demonstrate that our method can generate realistic new actions easily for objects and achieve high-quality rendering.
title RigGS: Rigging of 3D Gaussians for Modeling Articulated Objects in Videos
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.16822