AI Paper Atlas

Research map / 3D vision / Gaussian splatting

Hierarchy UGP: Hierarchy Unified Gaussian Primitive for Large-Scale Dynamic Scene Reconstruction

Hongyang Sun, Qinglin Yang, Jiawei Wang, Zhen Xu, Chen Liu, Yida Wang, Kun Zhan, Hujun Bao, Xiaowei Zhou, Sida Peng

ICCV 2025 · 3D vision › Gaussian splatting · approach: Monocular dynamic scene reconstruction · 0 citations (OpenAlex)

Read the paper

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 Block-wise Objects training (ablation)Hierarchy UGP · Waymo vehicle regionsPSNR+: 21.51 -> 23.15Improved
add Block-wise Objects training (ablation)Hierarchy UGP · Waymo overallPSNR: 28.58 -> 28.64No clear effect
add temporal scale initialization (ablation)Hierarchy UGP · Waymo human regionsPSNR: 23.22 -> 34.93Improved
add temporal scale initialization (ablation)Hierarchy UGP · Waymo human regionsSSIM: 0.791 -> 0.943Improved
add virtual warping view supervision (ablation)Hierarchy UGP · Dynamic City lane shift 2m NVSvisual qualityImproved
use Hierarchy UGP instead of Hierarchical GSGaussian scene model · Dynamic City Large 001PSNR: 26.74 -> 27.01Improved
use Hierarchy UGP instead of OmniRedynamic scene model · Dynamic City Large 002PSNR: 22.17 -> 25.25Improved
use Hierarchy UGP instead of 4D-GSdynamic scene model · Dynamic City Sub-scene 002PSNR: 25.51 -> 27.61Improved
use Hierarchy UGP instead of Street GS for extrapolationdynamic scene model · Dynamic City Large 001 lane shiftFID: 90.99 -> 68.27Improved
use Hierarchy UGP instead of Street GSdynamic scene model · Waymo scene 023 reconstruction pedestrianPSNR*: 24.75 -> 34.10Improved
use Hierarchy UGP instead of Street GS for novel view synthesisdynamic scene model · Waymo scene 023 NVS pedestrianPSNR*: 21.79 -> 21.64Hurt
use Hierarchy UGP instead of OmniRe for novel view synthesisdynamic scene model · Waymo scene 023 NVS pedestrianPSNR*: 23.07 -> 21.64Hurt

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