AI Paper Atlas

Research map / 3D vision / Gaussian splatting

ViewSplat: View-Adaptive Dynamic Gaussian Splatting for Feed-Forward Synthesis

Moonyeon Jeong, Seunggi Min, Suhyeon Lee, Hongje Seong

ECCV 2026 · 3D vision › Gaussian splatting · approach: Sparse-view Gaussian splatting for novel views

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 ViewSplat view-adaptive refinementSPFSplat · RE10K (2 views)PSNR: 25.484 -> 26.317Improved
add ViewSplat view-adaptive refinementSPFSplatV2-L · RE10K (2 views)PSNR: 25.668 -> 26.798Improved
add ViewSplat view-adaptive refinementSPFSplatV2-L · ACID (2 views)PSNR: 26.674 -> 27.509Improved
add ViewSplat view-adaptive refinement (zero-shot)SPFSplatV2-L · DL3DV (zero-shot from RE10K)PSNR: 19.743 -> 20.563Improved
add ViewSplat view-adaptive refinementSPFSplat · RE10K (10 context views)PSNR: 27.159 -> 29.010Improved
add ViewSplat view-adaptive refinementSPFSplat · RE10K (5 context views)LPIPS: 0.122 -> 0.124No clear effect
add ViewSplat on YoNoSplat backboneYoNoSplat · RE10K (6 views)PSNR: 25.071 -> 25.975Improved
add ViewSplat rendering overhead (target-pose conditioning)SPFSplat · RE10K rendering speedFPS: 386 -> 154Hurt
raise SH degree of baseline from 4 to 8 instead of using ViewSplat (ablation)SPFSplat · RE10KPSNR: 25.486 -> 25.418No clear effect
use hypernetwork View MLP instead of direct regression (ablation)ViewSplat · RE10Krendering speed (FPS): 72 -> 154Improved
reduce View MLP hidden dimension from 16 to 8 (ablation)ViewSplat · RE10KFPS: 154 -> 186Improved
reduce View MLP hidden dimension from 16 to 4 (ablation)ViewSplat · RE10KPSNR: 26.317 -> 26.302Hurt

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