AI Paper Atlas

Research map / Video / Video generation

Controlling Motion Transfer in Diffusion Transformers via Attention Heads

Sunyoung Jung, Jiwoo Park, Yoonseok Choi, Kyobin Choo, Ming-Hsuan Yang, Seong Jae Hwang

ECCV 2026 · Video › Video generation · approach: Efficient video diffusion models · 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
add selective structural head injection (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 61.3 -> 66.2Improved
add semantic-aware motion guidance (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 61.8 -> 66.2Improved
add both components to displacement optimization (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 59.6 -> 66.2Improved
add both components to displacement optimization (ablation)HALO · Motion transfer benchmarkTemporal Consistency: 89.5 -> 87.5Hurt
use motion-specific heads instead of all heads (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 63.1 -> 66.2Improved
use motion-specific heads instead of spatial heads (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 53.6 -> 66.2Improved
inject low-entropy heads (tau<7) rather than high-entropy (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 54.1 -> 66.2Improved
use top-k 4 instead of 10 (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 45.2 -> 66.2Improved
use semantic bias beta 0.1 instead of 0.3 (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 53.6 -> 66.2Improved
use 12 rather than 10 optimization steps (ablation)HALO · Motion transfer benchmarkMotion Fidelity: 50.1 -> 66.2Improved
HALO vs GWTF on Movie Scene DatasetHALO · Movie Scene DatasetMotion Fidelity: 46.6 -> 52.6Improved
HALO vs GWTF on hard motionHALO · Motion transfer benchmark (Hard motion)Motion Fidelity: 56.0 -> 59.2Improved

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