AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Continuous Adversarial Flow Models

Shanchuan Lin, Ceyuan Yang, Zhijie Lin, Hao Chen, Haoqi Fan

ECCV 2026 · Image generation › Fast diffusion sampling · approach: Rectified flow straightening · 5 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
post-train with the CAFM objectiveSiT-XL/2 · ImageNet 256, no guidanceFID: 8.26 -> 3.63Improved
continue training with the FM objective for 10 epochs (control) (ablation)SiT-XL/2 · ImageNet 256, no guidanceFID: 8.26 -> 8.64No clear effect
post-train with the CAFM objectiveJiT-H/16 · ImageNet 256, no guidanceFID: 7.17 -> 3.57Improved
continue training with the FM objective for 10 epochs (control) (ablation)JiT-H/16 · ImageNet 256, no guidanceFID: 7.17 -> 9.30Hurt
post-train with CAFM, best CFGSiT-XL/2 · ImageNet 256, guidedFID: 2.06 -> 1.53Improved
post-train with CAFM, best CFGJiT-H/16 · ImageNet 256, guidedFID: 1.86 -> 1.80Improved
post-train with CAFM at high guidance scale CFG 1.6SiT-XL/2 · ImageNet 256, CFG 1.6FID: 2.25 -> 2.37Hurt
post-train Z-Image with CAFM instead of FM (no PE, no CFG)Z-Image · GenEval 512pxoverall: 0.33 -> 0.44Improved
post-train Z-Image with CAFM instead of FM (no CFG)Z-Image · DPG-Bench 512pxoverall: 72.25 -> 77.21Improved
post-train Z-Image with CAFM instead of FM (with CFG)Z-Image · DPG-Bench 512pxoverall: 83.67 -> 85.21Improved
use N=16 discriminator updates per generator update instead of N=8 (ablation)SiT-XL/2 CAFM · ImageNet 256, no guidanceFID: 294.91 -> 3.63Improved
use N=16 instead of N=32 (ablation)SiT-XL/2 CAFM · ImageNet 256, no guidanceFID: 3.68 -> 3.63No clear effect

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