AI Paper Atlas

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

Exploring Multimodal Diffusion Transformers for Enhanced Prompt-based Image Editing

Joonghyuk Shin, Alchan Hwang, Yujin Kim, Daneul Kim, Jaesik Park

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 11 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
edit via image-projection replacement instead of fixed-seed regenerationMM-DiT editing (SD3-M) · PARTI 60 promptsLPIPS: 0.594 -> 0.380Improved
edit via image-projection replacement instead of fixed-seed regenerationMM-DiT editing (Flux.1-schnell) · PARTI 60 promptsLPIPS: 0.508 -> 0.279Improved
edit via image-projection replacement instead of prompt change at 20% timestepsMM-DiT editing (Flux.1-dev) · PARTI 60 promptsCLIP score: 0.311 -> 0.339Improved
edit via image-projection replacement instead of prompt change at 20% timestepsMM-DiT editing (Flux.1-schnell) · PARTI 60 promptsCLIP score: 0.289 -> 0.333Improved
edit via image-projection replacement instead of prompt changeMM-DiT editing (SD3-M) · PARTI 60 promptsLPIPS: 0.325 -> 0.380Hurt
replace only q_i,k_i image projections instead of full attention map (ablation)MM-DiT editing (Flux.1-dev) · text-misaligned prompt pairsedit stabilityImproved
use I2I block replacement instead of q_i,k_i projection replacement (ablation)MM-DiT editing · computation timespeed: up to 3xHurt
add top-5 block selection for attention masks (ablation)MM-DiT local blending (Flux.1-dev) · local editingmask noiseImproved
add Gaussian smoothing to attention masks (ablation)MM-DiT local blending · block-wise BCE, Soft mIoU, MSEmask qualityImproved
use T2I blocks instead of I2T blocks for masks (ablation)MM-DiT local blending · local editingmask qualityImproved
add local blending with top-5 block masks (ablation)MM-DiT editing (Flux.1-dev) · inference timeseconds: 55.9s vs. 53.7sNo clear effect
raise masking threshold theta from 0.2 to 0.5 (ablation)Ours with RF inversion (Flux.1-dev) · PIE-BenchPSNR: 19.92 -> 24.79Improved

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