AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DuoFlow: JVP-Free Finite-Difference Mean Flows for One-Step Image Generation

Zeming Li, Xiangyu Zhang, Ping Tan, Heung-Yeung Shum

ECCV 2026 · Image generation › Fast diffusion sampling · approach: Distillation for few-step generation · 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
replace JVP-based MeanFlow with DuoFlow finite-difference trainingMeanFlow-XL/2 · ImageNet 256x256 (1-NFE)FID: 3.43 -> 2.48Improved
replace JVP with one-forward differential estimatorderivative evaluation · SiT-XL/2 micro-benchmarkspeedup: 7.17x to 34.83x fasterImproved
DuoFlow trainer instead of MeanFlow JVP trainerXL/2 training · end-to-end wall-clockseconds per iteration: 0.89 -> 0.35 s/iterImproved
add sign randomization to one-sided differencing (ablation)DuoFlow · ImageNet 256x256 (80-epoch ablation)FID: 19.06 -> 14.11Improved
add interval-conditioned joint velocity estimation (ablation)DuoFlow · ImageNet 256x256 (80-epoch ablation)FID: 14.11 -> 11.30Improved
add progressive self-bootstrapping (ablation)DuoFlow · ImageNet 256x256 (80-epoch ablation)FID: 11.30 -> 9.17Improved
signed differencing alone vs JVP MeanFlow baseline (ablation)MeanFlow · ImageNet 256x256 (80-epoch ablation)FID: 14.13 -> 14.11No clear effect
use fixed kappa=0.9 instead of progressive alignment-gated kappa (ablation)DuoFlow self-bootstrapping · ImageNet 256x256 (80-epoch ablation)best FID: 9.17 -> 35.33Hurt
use progressive kappa instead of fixed 0.5-0.7 (ablation)DuoFlow self-bootstrapping · ImageNet 256x256 (80-epoch ablation)best FID: 9.36 -> 9.17Improved
use extreme step size (0.25x or 50x base) (ablation)DuoFlow differencing · ImageNet 256x256 (80-epoch ablation)FID: 9.17 -> 16.52 (0.25x), 14.21 (50x)Hurt
vary base step size from 1x to 5x or 10x (ablation)DuoFlow differencing · ImageNet 256x256 (80-epoch ablation)FID: 9.17 -> 9.12 (5x), 9.16 (10x)No clear effect
add RMS-based step rescaling (ablation)DuoFlow differencing · ImageNet 256x256 (80-epoch ablation)FID: 9.68 -> 9.17Improved

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