AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Geometry-Aware Dataset Condensation for Diffusion Model Training

Xiao Cui, Yulei Qin, Mo Zhu, Wengang Zhou, Hongsheng Li, Houqiang Li

ICML 2026 · Image generation › Fast diffusion sampling · approach: Diffusion models as perception tools · 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 swap-based refinement (Stage II) after greedy selection (ablation)geometry-aware selection · ImageNet 256x256 50K budgetFID-50K: 12.87 -> 11.01Improved
add swap-based refinement (Stage II) (ablation)geometry-aware selection · ImageNet 256x256 10K budgetFID-50K: 3.82 -> 3.43Improved
add optimal transport loss L_OT (ablation)geometry-aware selection · ImageNet 256x256 10K budgetFID-50K: 3.82 -> 3.43Improved
add mean-variance regularization L_sta (ablation)geometry-aware selection · ImageNet 256x256 10K budgetFID-50K: 4.62 -> 3.43Improved
add confidence regularization L_conf (ablation)geometry-aware selection · ImageNet 256x256 10K budgetIS: 337.0 -> 414.3Improved
use one-sided partial OT instead of balanced OT (ablation)geometry-aware selection · ImageNet 256x256 10K budgetFID-50K: 3.54 -> 3.43Improved
increase capacity factor kappa from 1.05 to 1.10 (ablation)one-sided partial OT · ImageNet 256x256 10K budgetFID-50K: 3.43 -> 3.58Hurt
vary dummy-source parameter gamma from 0.3 to 2.0 (ablation)one-sided partial OT · ImageNet 256x256 10K budgetFID-50K: 3.43 -> 3.47No clear effect
replace difficulty-ranking selection (D2C) with OT-based distribution alignmentdataset condensation for DiT-L/2 · ImageNet 512x512 10K budgetFID: 14.8 -> 6.17Improved
replace D2C with geometry-aware selection on SiT-L/2dataset condensation for SiT-L/2 · ImageNet 256x256 100K budgetFID-50K: 15.01 -> 8.83Improved
replace D2C with geometry-aware selection, longer trainingDiT-L/2 · ImageNet 256x256 50K budget, 300K itersFID-50K: 4.76 -> 3.34Improved
geometry-aware selection vs D2C at 10K budget on SiT-L/2 (256x256)SiT-L/2 · ImageNet 256x256 10K budgetFID-50K: 3.98 -> 3.56Improved

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