AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

ResCa: Residual Caching for Diffusion Transformers Acceleration

Haipeng Fang, Yu Li, Fan Tang, Yixing Lu, Juan Cao, Sheng Tang

CVPR 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration

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 trajectory-based clustering instead of feature-based clustering (ablation)ResCa · FLUX (N=8)image qualityImproved
use K=16 or K=32 clusters (ablation)ResCa · FLUX (N=8, O=1)image quality vs costImproved
raise residual order from 1 to 2 (ablation)ResCa-IT · DiT-XL/2 ImageNetFID: 2.62 -> 2.57Improved
raise residual order from 2 to 3 or 4 (ablation)ResCa-IT · DiT-XL/2 ImageNetFID: 2.57 -> 2.58No clear effect
use implicit Euler instead of ClusCa mixed first-order predictionResCa-IE · DiT-XL/2 ImageNet (N=5)FID: 2.65 -> 2.62Improved
use implicit Euler instead of ClusCaResCa-IE · FLUX DrawBench (N=5)Image Reward: 0.9825 -> 0.9958Improved
use implicit Euler instead of TaylorSeer forecastingResCa-IE · DiT-XL/2 ImageNetFID: 2.49 at 3.23x vs 2.49 at 2.77xImproved
use BDF2 implicit updateResCa-IB · FLUX DrawBench (N=7)Image Reward: 0.9737 -> 0.9889Improved

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