AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Sparse4DGS: 4D Gaussian Splatting for Sparse-Frame Dynamic Scene Reconstruction

Changyue Shi, Chuxiao Yang, Xinyuan Hu, Minghao Chen, Wenwen Pan, Yan Yang, Jiajun Ding, Zhou Yu, Jun Yu

AAAI 2026 · 3D vision › Gaussian splatting · approach: Deformable 4D Gaussian splatting · 11 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 Texture-Aware Deformation Regularization (TADR) (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.89 -> 22.34Improved
add Texture-Aware Canonical Optimization (TACO) (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.33 -> 22.34Improved
add eps_tex noise term in TACO (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.57 -> 22.34Improved
add eps_o noise term in TACO (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.81 -> 22.34Improved
use PCC loss instead of L1 for Ltex (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.71 -> 22.34Improved
use PCC loss instead of L1 for Ltadr (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 22.09 -> 22.34Improved
use texture-aware depth loss instead of plain depth PCC (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 21.46 -> 22.34Improved
raise Ltex weight from 0.01 to 1 (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 22.34 -> 20.17Hurt
raise Ltadr weight from 0.01 to 1 (ablation)Sparse4DGS · NeRF-DS (20 frames)PSNR: 22.34 -> 21.86Hurt
use Sparse4DGS vs Deform-3DGS with 5 FPS inputSparse4DGS · iPhone-4D (5 FPS input)PSNR: 21.12 -> 27.51Improved
use Sparse4DGS vs 4DGaussians on NeRF-Synthetic 20 framesSparse4DGS · NeRF-Synthetic (20 frames)PSNR: 22.47 -> 25.31Improved

Earlier methods this paper compares with

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