Fast Converging 3D Gaussian Splatting for 1-Minute Reconstruction

TL;DR

This work addresses real-time 3D reconstruction with 3D Gaussian Splatting under a strict 1-minute budget. It introduces a two-phase pipeline: a first round using noisy SLAM poses with forward/backward acceleration, anchor-based Neural-Gaussians for rapid convergence, and a global pose refinement; and a final round with accurate COLMAP poses that disables pose refinement and reverts to standard 3DGS, augmented by depth supervision and multi-view guided densification. Key contributions include SnugBox-based forward tiling with load balancing, per-Gaussian backward propagation, Neural-Gaussians to reduce parameters, and multi-view densification to accelerate convergence while preserving fidelity. The method achieves top performance on the SIGGRAPH Asia 3DGS Fast Reconstruction Challenge with PSNR around 28.4 dB, demonstrating fast convergence without sacrificing detail in photorealistic 3D reconstructions.

Abstract

We present a fast 3DGS reconstruction pipeline designed to converge within one minute, developed for the SIGGRAPH Asia 3DGS Fast Reconstruction Challenge. The challenge consists of an initial round using SLAM-generated camera poses (with noisy trajectories) and a final round using COLMAP poses (highly accurate). To robustly handle these heterogeneous settings, we develop a two-stage solution. In the first round, we use reverse per-Gaussian parallel optimization and compact forward splatting based on Taming-GS and Speedy-splat, load-balanced tiling, an anchor-based Neural-Gaussian representation enabling rapid convergence with fewer learnable parameters, initialization from monocular depth and partially from feed-forward 3DGS models, and a global pose refinement module for noisy SLAM trajectories. In the final round, the accurate COLMAP poses change the optimization landscape; we disable pose refinement, revert from Neural-Gaussians back to standard 3DGS to eliminate MLP inference overhead, introduce multi-view consistency-guided Gaussian splitting inspired by Fast-GS, and introduce a depth estimator to supervise the rendered depth. Together, these techniques enable high-fidelity reconstruction under a strict one-minute budget. Our method achieved the top performance with a PSNR of 28.43 and ranked first in the competition.

Figures (6)

• Figure 1: The illustration of the compact bounding strategy, adapted from HansonSpeedy. The blue thin boxes denote the bounding boxes, and the yellow regions indicate the tiles being written. The first column shows the original 3DGS. The second column our method. The third column presents the pruned bounding box after applying our compact tiling strategy.
• Figure 2: Illustration of sequential writing (left) versus load-balanced writing (right). The blue boxes denote the initially recorded tiles, while the red regions indicate the tiles that are finally written.
• Figure 3: Illustration of per-pixel parallelism (left) and per-Gaussian parallelism (right). In per-pixel parallelism, each lane is responsible for processing one pixel, whereas in per-Gaussian parallelism, each lane handles one Gaussian splat.
• Figure 4: Comparison between using Scaffold-GS as the scene representation (left) and the original 3DGS formulation (right). The Neural-Gaussian representation leads to faster convergence and captures finer scene details.
• Figure 5: Comparison between non-densified initial point clouds (left) and densified initial point clouds (right). Densifying the initialization introduces more geometric details and improves scene coverage.