BlitzGS: City-Scale Gaussian Splatting at Lightning Speed

2026-05-13

AI Summary

This work addresses the high computational cost and slow training speed of city-scale 3D Gaussian splatting by proposing a distributed framework that jointly optimizes system, model, and view levels. The approach employs an index-parity-based Gaussian partitioning strategy to enable distributed GPU memory storage, integrates single-pass cross-GPU rendering with importance-weighted density control, and introduces a dual culling mechanism based on both distance and contribution metrics. These innovations collectively reduce the number of active Gaussians and communication redundancy. Experimental results demonstrate that the method achieves an order-of-magnitude acceleration in training while preserving reconstruction quality, enabling efficient city-scale scene reconstruction within tens of minutes.

Abstract

We present BlitzGS, a distributed 3DGS framework that reduces active Gaussian workload for fast city-scale reconstruction. BlitzGS manages this workload at three coupled levels. At the system level, the framework shards Gaussians across GPUs by index parity rather than spatial blocks. This approach mitigates the cross-block visibility redundancy inherent in spatial partitioning. Furthermore, it distributes each rendering step through a single cross-GPU exchange that routes projected Gaussians to their tile owners. At the model level, scheduled importance-scoring passes shrink the global Gaussian population. During these passes, the framework generates a per-Gaussian visibility weight to bias density-control updates toward contributing primitives and a per-view importance mask for the view-level renderer. At the view level, BlitzGS trims each camera's active set with a distance-based LOD gate to exclude excessively fine primitives for the current frustum and the importance-based culling mask to skip Gaussians with negligible cross-view contribution. On large-scale benchmarks, BlitzGS matches the rendering quality of recent large-scale baselines while delivering an order-of-magnitude speedup, training city-scale scenes in tens of minutes. Our code is available at https: //github.com/AkierRaee/BlitzGS.

Problem

• city-scale reconstruction
• Gaussian Splatting
• large-scale 3D reconstruction
• rendering acceleration
• distributed rendering

Innovation

• Gaussian Splatting
• distributed rendering
• importance-based culling
• city-scale reconstruction
• level-of-detail (LOD)