GART: Gaussian Articulated Template Models
This provides an efficient representation for 3D reconstruction and rendering in computer vision, though it appears incremental as it builds on existing template models like SMPL.
The paper tackles the problem of capturing and rendering non-rigid articulated subjects from monocular videos by introducing GART, which uses moving 3D Gaussians and a template model prior, achieving reconstruction in seconds to minutes and rendering at over 150fps.
We introduce Gaussian Articulated Template Model GART, an explicit, efficient, and expressive representation for non-rigid articulated subject capturing and rendering from monocular videos. GART utilizes a mixture of moving 3D Gaussians to explicitly approximate a deformable subject's geometry and appearance. It takes advantage of a categorical template model prior (SMPL, SMAL, etc.) with learnable forward skinning while further generalizing to more complex non-rigid deformations with novel latent bones. GART can be reconstructed via differentiable rendering from monocular videos in seconds or minutes and rendered in novel poses faster than 150fps.