AI Paper Atlas

Research map / Video / Video generation

EffiVMT: Video Motion Transfer via Efficient Spatial-Temporal Decoupled Finetuning

Yue Ma, Yulong Liu, Qiyuan Zhu, Xiangpeng Yang, Kunyu Feng, Xinhua Zhang, Zexuan Yan, Zhifeng Li, Sirui Han, Chenyang Qi, Qifeng Chen

ICLR 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 spatial-temporal decoupled LoRA (ablation)EffiVMT · MotionBenchMotion fidelity: 0.546 -> 0.971Improved
add adaptive RoPE (ablation)EffiVMT · MotionBenchMotion fidelity: 0.655 -> 0.971Improved
add sparse motion sampling (ablation)EffiVMT · MotionBenchtuning time (s): 2068 -> 727Improved
add sparse motion sampling (ablation)EffiVMT · MotionBenchMotion fidelity: 0.975 -> 0.971No clear effect
add adaptive RoPE (ablation)EffiVMT · MotionBenchTemporal consistency: 0.817 -> 0.976Improved
add spatial-temporal decoupled LoRA (ablation)EffiVMT · MotionBenchtuning time (s): 971 -> 727Hurt
classify heads with pseudo-label attention maps instead of random assignment (ablation)EffiVMT · MotionBenchMotion fidelity: 0.922 -> 0.971Improved
use alpha 1.25 instead of 0.75 for head classification (ablation)EffiVMT · MotionBenchMotion fidelity: 0.925 -> 0.971Improved
use alpha 1.25 instead of 1.75 (ablation)EffiVMT · MotionBenchMotion fidelity: 0.957 -> 0.971Improved
use EffiVMT instead of MotionDirector (WAN-2.1 re-implementation)EffiVMT · MotionBench, Wan2.1 backbonetuning time (s): 4641 -> 727Improved
switch backbone from CogVideoX to Wan2.1EffiVMT · MotionBenchMotion fidelity gain over MotionDirector: 0.016 -> 0.04Improved

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