AI Paper Atlas

Research map / 3D vision / Gaussian splatting

COSMOS: Coherent SuperGaussian Modeling with Spatial Priors for Sparse-View 3D Splatting

Chaeyoung Jeong, Kwangsu Kim

AAAI 2026 · 3D vision › Gaussian splatting · approach: Regularization and primitive design for 3DGS · 0 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
add supergaussian self-attention (ablation)COSMOS · Blender (3 views)PSNR: 18.74 -> 18.96Improved
add supergaussian self-attention (ablation)COSMOS · Blender (3 views)SSIM: 0.77 -> 0.79Improved
add supergaussian position regularization (ablation)COSMOS · Blender (3 views)PSNR: 18.24 -> 18.96Improved
add supergaussian position regularization (ablation)COSMOS · Blender (3 views)LPIPS: 0.21 -> 0.19Improved
add both supergaussian regularization and self-attention (ablation)COSMOS · Blender (3 views)PSNR: 18.22 -> 18.96Improved
use COSMOS instead of SplatFields (its base)COSMOS · Blender (3 views)PSNR: 18.46 -> 18.96Improved
use COSMOS instead of FreeNeRF with more viewsCOSMOS · Blender (6 views)PSNR: 23.67 -> 22.78Hurt
use COSMOS instead of SplatFieldsCOSMOS · DTU (3 views)PSNR: 21.07 -> 21.86Improved
use COSMOS instead of DNGaussianCOSMOS · DTU (3 views)LPIPS: 0.15 -> 0.23Hurt
use COSMOS instead of MVPGS for depthCOSMOS · DTU depth (3 views)depth MAE: 0.28 -> 0.22Improved
use COSMOS instead of MVPGS for depthCOSMOS · DTU depth (3 views)SROCC: 0.44 -> 0.41Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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