AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

AccelAes: Accelerating Diffusion Transformers for Training-Free Aesthetic-Enhanced Image Generation

Xuanhua Yin, CHUANZHI XU, Haoxian Zhou, Boyu Wei, Weidong Cai

ECCV 2026 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis · 4 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 sparse computation to step-level cache (ablation)AccelAes (Lumina-Next) · Pick-a-PicImageReward: 0.743 -> 0.841Improved
add step-level cache alone (ablation)dense Lumina-Next · Pick-a-PicImageReward: 0.752 -> 0.743Hurt
add spatial allocation alone (ablation)dense Lumina-Next · Pick-a-PicImageReward: 0.752 -> 0.818Improved
add AesMask instead of uniform mask (ablation)AccelAes (Lumina-Next) · Pick-a-PicImageReward: 0.8222 -> 0.8410Improved
use aesthetic anchor tokens instead of non-aesthetic tokens (ablation)AesMask · Pick-a-PicImageReward: 0.8207 -> 0.8410Improved
add spatial FFN sparsity on top of spatial attention sparsity (ablation)spatial-only variant · Pick-a-PicImageReward: 0.818 -> 0.825No clear effect
raise skip_ratio from 0.50 to 0.70 (ablation)AccelAes · Pick-a-PicImageReward: 0.8410 -> 0.7643Hurt
lower skip_ratio from 0.50 to 0.40 (ablation)AccelAes · Pick-a-PicImageReward: 0.8410 -> 0.8334No clear effect
delay mask_step from 5 to 15 (ablation)AccelAes · Pick-a-Picspeedup: 2.11x -> 1.42xHurt
use static one-shot mask instead of per-step refresh (ablation)AccelAes · hard non-centric promptsImageReward: 0.381 -> 0.407Improved
use static one-shot mask instead of per-step refresh (ablation)AccelAes · hard non-centric promptsspeedup: 1.82x -> 2.08xImproved
add Spatial-CFG (ablation)AccelAes · blind human studyoverall preference win rate: 66%Improved

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