AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Improving Sparse-View 3DGS Generalization via Flat Minima Optimization

Kangmin Seo, Sangeek Hyun, MinKyu Lee, Jae-Pil Heo

ECCV 2026 · 3D vision › Gaussian splatting · approach: Regularization and primitive design for 3DGS · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use anisotropic scale-adaptive noise instead of isotropic noise (ablation)flat-minima 3DGS · LLFF 3-viewPSNR: 20.54 -> 20.88Improved
perturb Gaussian position rather than scale, rotation or opacity (ablation)flat-minima 3DGS · LLFF 3-viewPSNR: 20.21-20.23 -> 20.88Improved
also perturb scale in addition to position (ablation)flat-minima 3DGS · LLFF 3-viewPSNR: 20.88 -> 20.33Hurt
apply perturbation stochastically per Gaussian instead of mixed loss (ablation)flat-minima 3DGS · LLFF 3-viewLPIPS: 0.212 -> 0.184Improved
add linear perturbation magnitude scheduling (ablation)flat-minima 3DGS · LLFF 3-viewPSNR: 20.69 -> 20.88Improved
add periodic reinitialization of scale, rotation, SH (ablation)flat-minima 3DGS · LLFF 3-viewPSNR: 20.58 -> 20.88Improved
add flat-minima plug-in to FSGSFSGS · LLFF 3-viewPSNR: 20.43 -> 21.03Improved
add flat-minima plug-in to DropGaussianDropGaussian · LLFF 3-viewPSNR: 20.76 -> 21.05Improved
add flat-minima plug-in to Difix3D+Difix3D+ · Mip-NeRF360 12-viewPSNR/LPIPS: PSNR 19.25 -> 19.63; LPIPS 0.375 -> 0.386Mixed
replace 3DGS post-optimization with ours after AnySplatAnySplat · Mip-NeRF360 12-viewPSNR: 16.97 -> 17.36Improved
apply method with more input views (36-view)3DGS · Mip-NeRF360 36-viewPSNR: 25.15 -> 25.01Hurt
apply method vs DropGaussian on Mip-NeRF360 12-view3DGS · Mip-NeRF360 12-viewPSNR: 19.74 -> 19.50 PSNR; LPIPS 0.364 -> 0.348Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy