QuickSplat: Fast 3D Surface Reconstruction via Learned Gaussian Initialization
This addresses the problem of efficient and accurate 3D modeling for applications like mixed reality and robotics, representing an incremental improvement over existing methods.
The paper tackles the slow optimization and geometry issues in 3D surface reconstruction by introducing QuickSplat, which uses learned priors to initialize Gaussian splatting, accelerating runtime by 8x and reducing depth errors by up to 48% compared to state-of-the-art methods.
Surface reconstruction is fundamental to computer vision and graphics, enabling applications in 3D modeling, mixed reality, robotics, and more. Existing approaches based on volumetric rendering obtain promising results, but optimize on a per-scene basis, resulting in a slow optimization that can struggle to model under-observed or textureless regions. We introduce QuickSplat, which learns data-driven priors to generate dense initializations for 2D gaussian splatting optimization of large-scale indoor scenes. This provides a strong starting point for the reconstruction, which accelerates the convergence of the optimization and improves the geometry of flat wall structures. We further learn to jointly estimate the densification and update of the scene parameters during each iteration; our proposed densifier network predicts new Gaussians based on the rendering gradients of existing ones, removing the needs of heuristics for densification. Extensive experiments on large-scale indoor scene reconstruction demonstrate the superiority of our data-driven optimization. Concretely, we accelerate runtime by 8x, while decreasing depth errors by up to 48% in comparison to state of the art methods.