AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

DynFusion: Rethinking Condition Fusion for Adaptive Multi-Conditional Text-to-Image Generation

Zheng Fang, Lichuan Xiang, Xu Cai, Bing Wang, Bo Yang, Hongkai Wen

CVPR 2026 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 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
use adaptive dynamic condition selection instead of uniformly activating all conditions (ablation)DynFusion · Subject-InsertionFID: 5.06 -> 4.53Improved
use adaptive selection instead of uniform: compute (ablation)DynFusion · Subject-InsertionFLOPs: 15.10T -> 7.76TImproved
sigmoid free selection instead of single-condition softmax selection (ablation)DynFusion · Subject-InsertionFID: 4.89 -> 4.53Improved
use decoupled multi-modal attention instead of MMA (ablation)DynFusion · Subject-InsertionFID: 5.19 -> 4.53Improved
use DMMA instead of CMMA (ablation)DynFusion · Subject-InsertionSSIM: 0.81 -> 0.80 SSIM; FID 4.75 -> 4.53Mixed
add Fusion-LoRA module (ablation)DynFusion · Subject-InsertionFID: 5.93 -> 4.53Improved
set condition sparsity to 50% vs 100% (uniform) (ablation)DynFusion · Subject-InsertionFID: 5.06 -> 4.77Improved
reduce condition sparsity to 30% (ablation)DynFusion · Subject-InsertionFID: 4.77 -> 5.19Hurt
DynFusion vs UniCombine on multi-spatial generationDynFusion · Multi-SpatialFLOPs: 16.21T -> 8.27TImproved
DynFusion vs UniCombine on text consistencyDynFusion · Multi-SpatialCLIP-T: 33.48 -> 33.44No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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