AI Paper Atlas

Research map / Video / Video generation

Towards Decompositional Human Motion Generation with Energy-Based Diffusion Models

Jianrong Zhang, Hehe Fan, Yi Yang

CVPR 2026 · Video › Video generation · approach: Text-to-motion diffusion

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
finetune with DecompML-generated motions added (ablation)SALAD · HumanML3D testFID: 0.076 -> 0.060Improved
finetune with DecompML-generated motions added (ablation)SALAD · HumanML3D testTop-1: 0.581 -> 0.580No clear effect
finetune with DecompML-generated motions added (ablation)EnergyMoGen · HumanML3D testFID: 0.188 -> 0.147Improved
orthogonal self-supervised decomposition (OSS) trainingSALAD-style latent diffusion · HumanML3D testR-Precision Top-1: 0.581 -> 0.588Improved
compositional training (DEMOGEN-OSS latent)SALAD · HumanML3D testFID: 0.076 -> 0.092No clear effect
semantic-aware vs latent-aware EBMDEMOGEN · HumanML3D testFID: EXP: 0.078 vs 0.116; OSS: 0.092 vs 0.104Mixed
DEMOGEN-EXP compositional trainingEnergyMoGen+SEF · MTT compositionalR@3: 28.0 -> 31.9Improved
DEMOGEN-EXP compositional trainingEnergyMoGen+SEF · MTT compositionalFID: 0.604 -> 0.621Hurt
DEMOGEN-EXP compositional trainingSALAD/EnergyMoGen · DecompML compositionTop-1: 0.342 -> 0.559Improved
DEMOGEN-OSS multi-concept trainingEMG+AGD · MTT multi-conceptFID: 0.587 -> 0.580Improved

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