AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

EasyTune: Efficient Step-Aware Fine-Tuning for Diffusion-Based Motion Generation

Xiaofeng Tan, Wanjiang Weng, Haodong Lei, Hongsong Wang

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Guidance for conditional diffusion · 3 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
use EasyTune step-wise reward optimization instead of trajectory-level DRaFT-50MLD fine-tuning · HumanML3DFID: 0.197 -> 0.132Improved
use EasyTune instead of DRaFT-50 for memoryMLD fine-tuning · HumanML3Dpeak memory (GB): 37.32 -> 22.10Improved
apply EasyTune to MLDMLD · HumanML3DFID: 0.473 -> 0.194Improved
apply EasyTune to MLD++MLD++ · HumanML3DFID: 0.073 -> 0.073No clear effect
apply EasyTune to MotionDiffuseMotionDiffuse · HumanML3DR-Precision Top-1: 0.491 -> 0.488Mixed
apply EasyTune to MoDiffuseMoDiffuse · KIT-MLDiversity: 11.10 -> 10.63Hurt
add KL regularization to EasyTune (ablation)EasyTune · HumanML3DFID: 0.132 -> 0.172Hurt
add KL regularization to EasyTune (ablation)EasyTune · HumanML3Dpeak memory (GB): 22.10 -> 29.70Hurt
optimize all denoising steps instead of only final 20 steps (ablation)EasyTune · HumanML3DFID: 0.184 -> 0.132Improved
optimize all denoising steps instead of only initial 20 steps (ablation)EasyTune · HumanML3DFID: 0.158 -> 0.132Improved
use linearly decreasing step weights emphasizing early steps (ablation)EasyTune · HumanML3DR-Precision Top-1: 0.581 -> 0.584No clear effect
add SPL self-refining preference learning to ReAlign reward model (ablation)ReAlign · HumanML3D text-motion retrievalR@1: 67.59 -> 69.31Improved

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