SkeletonGaussian: Editable 4D Generation through Gaussian Skeletonization

TL;DR

SkeletonGaussian introduces a hierarchical, skeleton-driven framework for editable 4D dynamic 3D object generation from monocular videos, improving interpretability and control over motion synthesis.

Contribution

It presents a novel hierarchical articulated representation that explicitly decomposes motion into rigid and non-rigid components, enabling more controllable and editable 4D generation.

Findings

Outperforms existing methods in generation quality

Enables intuitive motion editing

Provides a new paradigm for editable 4D synthesis

Abstract

4D generation has made remarkable progress in synthesizing dynamic 3D objects from input text, images, or videos. However, existing methods often represent motion as an implicit deformation field, which limits direct control and editability. To address this issue, we propose SkeletonGaussian, a novel framework for generating editable dynamic 3D Gaussians from monocular video input. Our approach introduces a hierarchical articulated representation that decomposes motion into sparse rigid motion explicitly driven by a skeleton and fine-grained non-rigid motion. Concretely, we extract a robust skeleton and drive rigid motion via linear blend skinning, followed by a hexplane-based refinement for non-rigid deformations, enhancing interpretability and editability. Experimental results demonstrate that SkeletonGaussian surpasses existing methods in generation quality while enabling intuitive motion editing, establishing a new paradigm for editable 4D generation. Project page: https://wusar.github.io/projects/skeletongaussian/