AI Paper Atlas

Research map / 3D vision / Gaussian splatting

F⁴Splat: Feed-Forward Predictive Densification for Feed-Forward 3D Gaussian Splatting

Injae Kim, Chaehyeon Kim, Minseong Bae, Minseok Joo, Hyunwoo Kim

ECCV 2026 · 3D vision › Gaussian splatting · approach: Compact and efficient Gaussian splatting

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 learned densification score instead of random allocation (ablation)F4Splat · RE10K (24 views, 20% Gaussian budget)PSNR: 24.68 -> 25.47Improved
add learned densification score instead of Sobel frequency heuristic (ablation)F4Splat · RE10K (24 views, 20% Gaussian budget)PSNR: 25.36 -> 25.47Improved
add level-wise Gaussian supervision during training (ablation)F4Splat · RE10K (24 views, 20% Gaussian budget)PSNR: 24.25 -> 25.47Improved
add scene-scale regularization loss (ablation)F4Splat · RE10K (24 views, 20% Gaussian budget)PSNR: 4.82 -> 25.47Improved
use novel-view training with aligned target camera and scene-scale reg instead of AnySplat-like input-view training with frozen VGGT distillation (ablation)F4Splat · VGGT-prior training comparisonPSNR: 20.59 -> 24.62Improved
same VGGT-prior strategy change on training time (ablation)F4Splat · VGGT-prior training comparisontraining time: 5H 35M -> 3H 03MImproved
use F4Splat tau- (447K Gaussians) instead of AnySplatF4Splat · RE10K (8 views)PSNR: 24.07 -> 25.73Improved
use F4Splat tau+ with far fewer Gaussians instead of AnySplatF4Splat · RE10K (24 views)PSNR: 25.40 -> 26.01Improved
use F4Splat instead of NoPoSplat when trained on RE10K+DL3DVF4Splat · DTU (2 views)PSNR: 17.51 -> 19.76Improved
use F4Splat tau- instead of VicaSplat uncalibratedF4Splat · ACID (2 views)PSNR: 24.55 -> 26.26Improved
add adaptive allocation over uniform allocation (compute cost)F4Splat · Computational overheadinference time: 0.440 -> 0.488Hurt
use F4Splat pose estimation over VGGTF4Splat · RE10K relative pose AUC@5AUC: 0.335 -> 0.541Improved

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