
This paper introduces SO(2)-Equivariant Gaussian Sculpting Networks (GSNs) as an efficient approach for 3D object reconstruction from single-view image observations.


coremsg

efficient-single-view-3d-reconstruction-with-so-2-equivariant-gaussian-sculpting-networks


Efficient Single-View 3D Reconstruction with SO(2)-Equivariant Gaussian Sculpting Networks


title_rewrite


A novel approach called DIG3D that marries Gaussian splatting with a deformable transformer to efficiently and accurately reconstruct 3D objects from a single RGB image.


efficient-3d-reconstruction-from-single-view-rgb-images-using-deformable-transformer-and-gaussian-splatting


Efficient 3D Reconstruction from Single-View RGB Images using Deformable Transformer and Gaussian Splatting
