AI Paper Atlas

Research map / 3D vision / Gaussian splatting

PointSplat: Compact Gaussian Splatting via Human-Centric Prediction

Yujie Guo, Yudong Jin, Lingteng Qiu, Zehong Shen, Zhen Xu, Zhang Jing, Xianchao Shen, Hujun Bao, Sida Peng, Xiaowei Zhou

ECCV 2026 · 3D vision › Gaussian splatting · approach: Animatable Gaussian human avatars · 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 voxel point encoding (ablation)PointSplat · DNA-RenderingPSNR: 26.60 -> 27.18Improved
add voxel point encoding (ablation)PointSplat · DNA-RenderingGS-Num: 160k -> 71kImproved
add ray casting (ablation)PointSplat · DNA-RenderingPSNR: 26.32 -> 27.18Improved
add ray casting (ablation)PointSplat · DNA-RenderingGS-Num: 224k -> 71kImproved
use alternating attention instead of full self-attention (ablation)PointSplat · DNA-RenderingPSNR and time: 27.09 -> 27.18; time 1.1 -> 0.4Improved
use visual hull instead of VGGT depth proxy (ablation)PointSplat · DNA-RenderingPSNR: 25.95 -> 27.18Improved
use visual hull instead of bounding box proxy (ablation)PointSplat · DNA-RenderingPSNR: 23.89 -> 27.18Improved
use visual hull instead of frustum sample proxy (ablation)PointSplat · DNA-RenderingPSNR: 22.92 -> 27.18Improved
reduce voxel size from 0.0050 to 0.0030 (ablation)PointSplat · DNA-RenderingPSNR and GS-Num: 27.18 -> 27.52; 71k -> 189kMixed
increase voxel size from 0.0050 to 0.0100 (ablation)PointSplat · DNA-RenderingPSNR: 27.18 -> 25.51Hurt
drop 10% of foreground mask pixels at inference (ablation)PointSplat · DNA-RenderingPSNR: 26.74 -> 24.55Hurt
use human-centric prediction instead of view-centric DepthSplatPointSplat · DNA-RenderingPSNR: 23.98 -> 27.18Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 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