AI Paper Atlas

Research map / 3D vision / 3D generation

TRACE: Learning 3D Gaussian Physical Dynamics from Multi-view Videos

Jinxi Li, Ziyang Song, Bo Yang

ICCV 2025 · 3D vision › 3D generation · approach: Learning physics from video · 5 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 auxiliary deformation field f_defo (ablation)TRACE · Dynamic MultipartPSNR: 19.206 -> 33.481Improved
derive future motion from physics parameters at t' rather than querying for each future time (ablation)TRACE · Dynamic MultipartPSNR: 29.602 -> 33.481Improved
use equivalent velocity/acceleration parametrization (ablation)TRACE · Dynamic MultipartPSNR: 29.986 -> 33.481Improved
use training time difference 2 delta t instead of delta t (ablation)TRACE · Dynamic MultipartPSNR: 32.852 -> 33.481Improved
use 2 delta t instead of 3 delta t (ablation)TRACE · Dynamic ObjectPSNR: 29.787 -> 31.511Improved
use 2nd-order instead of 1st-order dynamics (ablation)TRACE · Dynamic ObjectPSNR: 28.522 -> 31.511Improved
use 2nd-order instead of 3rd-order dynamics (ablation)TRACE · Dynamic Indoor ScenePSNR: 34.086 -> 34.824Improved
learn hard physics parameters instead of PINN soft constraints (vs DefGS_nvfi)TRACE · Dynamic MultipartPSNR: 28.455 -> 33.481Improved
use 4DGS as auxiliary deformation fieldTRACE · NVIDIA Dynamic ScenePSNR: 29.341 -> 22.554Hurt
replace baseline deformable 3DGS with TRACE dynamics in continual learningDefGS_nvfi · ParticleNeRFPSNR: 26.181 -> 27.544Improved
cluster learned physical parameters for motion segmentation vs NVFiNVFi · Dynamic Indoor SceneAP: 91.21 -> 95.82Improved

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