Research method that rigs animatable Gaussian characters with an inner kinematic skeleton and outer free-form bones.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
GaussiAnimate is a research project from Westlake University (Endless AI Lab, per the project page) on rigging animatable characters such as quadrupeds and clothed humans. It proposes a "Scaffold-Skin Rigging System" called skelebones, which combines an inner kinematic skeleton for intuitive control with outer free-form bones that capture non-rigid surface dynamics such as loose clothing and soft tissue.
The pipeline has three steps. Temporally consistent deformable Gaussians are compressed into free-form bones. A Mean Curvature Skeleton is extracted from the canonical Gaussians and refined over time. The skeleton and bones are then bound by non-parametric partwise motion matching, which retrieves and blends observed bone motions to synthesize new ones. The authors report 17.3% PSNR gains over Linear Blend Skinning and 21.7% over Bag-of-Bones on unseen poses, and say the motion-matching step also works on mesh representations with about 1000 frames of data.
It is evaluated on VTO, DeformingThings4D and ActorsHQ datasets. It needs a reconstructed, captured motion sequence as input. The project page states that code will be made publicly available for research purposes. Creator coverage says code has been released, but no repository link or license appeared in the provided material. There is no hosted product.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Animates 3D characters with soft outer layers (loose clothes, animals) by adding floating nodes on top of a skeleton to capture jiggle and flow. Requires a captured motion sequence; code is on GitHub for research use.
Animates 3D characters by combining skeleton motion with a flexible outer-surface representation, handling clothing, fur and animals. Code released.
Closest neighbours by meaning, not just by category.
Model
Research method that reconstructs a 4D Gaussian Splatting human from a single casual monocular video.
Model
Video-diffusion LoRA that repairs artifacts in 3DGS, NeRF, mesh and point-cloud renders while keeping the camera path.
Model
Text-prompted motion generation model that animates rigged 3D assets with arbitrary skeleton topologies, no retraining.
Research · Google
Research method for free-view rendering and relighting of full-body dynamic humans from four sparse RGB views.
Model
Feed-forward 3D Gaussian Splatting model that reconstructs scenes from arbitrary input views, with or without camera poses.
Research
Research method that reconstructs complete 4D (3D plus time) scenes from a single monocular video using Gaussian Splatting.
