AI Paper Atlas

Research map / Video / Video generation

MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model

Wenxun Dai, Ling-Hao Chen, Jingbo Wang, Jinpeng Liu, Bo Dai, Yansong Tang

ECCV 2024 · Video › Video generation · approach: Text-to-motion diffusion · 113 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
latent consistency distillation of MLD to 1-step inferenceMLD · HumanML3D text-to-motionAITS (s): 0.225 -> 0.030Improved
latent consistency distillation of MLD to 1-step inferenceMLD · HumanML3D text-to-motionFID: 0.450 -> 0.467No clear effect
increase MotionLCM inference steps from 1 to 4MotionLCM · HumanML3D text-to-motionFID: 0.467 -> 0.304Improved
dynamic training guidance scale [5,15] instead of static 7.5 (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.479 -> 0.467Improved
widen training guidance range to [2,18] (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.467 -> 0.481Hurt
use EMA rate 0.95 instead of 0 (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.505 -> 0.467Improved
use skipping interval k=20 instead of k=1 (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.635 -> 0.467Improved
increase skipping interval to k=50 (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.467 -> 0.547Hurt
use Huber loss instead of L2 loss (ablation)MotionLCM · HumanML3D text-to-motion (1-step)FID: 0.622 -> 0.467Improved
add motion ControlNet (ablation)MotionLCM · HumanML3D motion control (1-step)Loc. err.: 0.2302 -> 0.0344Improved
add motion-space control loss (lambda=1) (ablation)MotionLCM ControlNet · HumanML3D motion control (1-step)Loc. err.: 0.0344 -> 0.0147Improved
increase control loss weight to 10 (ablation)MotionLCM ControlNet · HumanML3D motion control (1-step)FID: FID 0.419 -> 0.636; Loc. err. 0.0147 -> 0.0097Mixed

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