AI Paper Atlas

Research map / Video / Video generation

What Moves? Localized Motion Representations for Compositional Scene Control

Frank Fundel, Malek Ben Alaya, Thomas Ressler-Antal, Stefan Andreas Baumann, Bjorn Ommer

ECCV 2026 · Video › Video generation · approach: Trajectory-guided motion control · 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
keep full-scene context instead of cropping the RGB input (ablation)Ours · A2D localized action classificationaccuracy: 27.98 -> 33.90Improved
use actor-specific mask instead of global full-frame mask (ablation)Ours · A2D localized action classificationaccuracy: 21.75 -> 33.90Improved
use localized motion encoder instead of DisMoOurs · A2D localized action classificationaccuracy: 27.24 -> 33.90Improved
use localized representation instead of DisMo for motion transferCogVideoX-5B · Controllable motion transfermotion selectivity: 0.1377 -> 0.2304Improved
use localized representation instead of DisMo for motion transferCogVideoX-5B · Controllable motion transferout-of-region motion leakage: 0.5604 -> 0.5158Improved
use localized representation instead of ATIWan 2.1-14B vs CogVideoX-5B · Controllable motion transferin-region motion fidelity: 0.8111 -> 0.7462Hurt
use localized representation instead of ATIWan 2.1-14B vs CogVideoX-5B · Composable motion transfermotion selectivity: 0.3291 -> 0.4114Improved
use localized representation instead of ATIWan 2.1-14B vs CogVideoX-5B · Composable motion transferin-region motion fidelity: 0.7148 -> 0.6718Hurt
use localized representation instead of DisMoglobal setting · SSv2 zero-shot classificationtop-1 accuracy: 24.6 -> 26.5Improved
use localized representation instead of DisMoglobal setting · Diving48 zero-shot classificationtop-1 accuracy: 20.9 -> 17.1Hurt
use localized representation instead of DisMoSemanticMoments synthetic retrieval · motion retrievaltop-1 accuracy: 90.0 vs 89.6No clear effect
use localized representation instead of global DisMoLAOM DCS action probing · mean NMSENMSE: 0.6415 -> 0.4339Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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