2D Gaussian Splatting for Geometrically Accurate Radiance Fields
This addresses the issue of multi-view inconsistent geometry in radiance field reconstruction for computer vision and graphics applications, representing an incremental improvement over 3D Gaussian Splatting.
The paper tackles the problem of inaccurate surface representation in 3D Gaussian Splatting for radiance fields by introducing 2D Gaussian Splatting, which collapses 3D volumes into 2D planar Gaussian disks to achieve geometrically accurate reconstructions with noise-free and detailed geometry while maintaining competitive appearance quality, fast training, and real-time rendering.
3D Gaussian Splatting (3DGS) has recently revolutionized radiance field reconstruction, achieving high quality novel view synthesis and fast rendering speed without baking. However, 3DGS fails to accurately represent surfaces due to the multi-view inconsistent nature of 3D Gaussians. We present 2D Gaussian Splatting (2DGS), a novel approach to model and reconstruct geometrically accurate radiance fields from multi-view images. Our key idea is to collapse the 3D volume into a set of 2D oriented planar Gaussian disks. Unlike 3D Gaussians, 2D Gaussians provide view-consistent geometry while modeling surfaces intrinsically. To accurately recover thin surfaces and achieve stable optimization, we introduce a perspective-correct 2D splatting process utilizing ray-splat intersection and rasterization. Additionally, we incorporate depth distortion and normal consistency terms to further enhance the quality of the reconstructions. We demonstrate that our differentiable renderer allows for noise-free and detailed geometry reconstruction while maintaining competitive appearance quality, fast training speed, and real-time rendering.