SkipGS: Post-Densification Backward Skipping for Efficient 3DGS Training

TL;DR

SkipGS introduces a view-adaptive backward skipping mechanism to significantly reduce training time in 3D Gaussian Splatting without sacrificing quality, by avoiding redundant gradient computations during post-densification.

Contribution

It proposes a novel view-adaptive backward gating method that selectively skips backpropagation in 3DGS training, enhancing efficiency without altering core components.

Findings

Reduces end-to-end training time by 23.1%.

Achieves a 42.0% reduction in post-densification time.

Maintains comparable reconstruction quality.

Abstract

3D Gaussian Splatting (3DGS) achieves real-time novel-view synthesis by optimizing millions of anisotropic Gaussians, yet its training remains expensive, with the backward pass dominating runtime in the post-densification refinement phase. We observe substantial update redundancy in this phase: many sampled views have near-plateaued losses and provide diminishing gradient benefits, but standard training still runs full backpropagation. We propose SkipGS with a novel view-adaptive backward gating mechanism for efficient post-densification training. SkipGS always performs the forward pass to update per-view loss statistics, and selectively skips backward passes when the sampled view's loss is consistent with its recent per-view baseline, while enforcing a minimum backward budget for stable optimization. On Mip-NeRF 360, compared to 3DGS, SkipGS reduces end-to-end training time by 23.1%, driven by a 42.0% reduction in post-densification time, with comparable reconstruction quality. Because it only changes when to backpropagate -- without modifying the renderer, representation, or loss -- SkipGS is plug-and-play and compatible with other complementary efficiency strategies for additive speedups.