AI Paper Atlas

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

LayerDiff: Exploring Text-guided Multi-layered Composable Image Synthesis via Layer-Collaborative Diffusion Model

Runhui Huang, Kaixin Cai, Jianhua Han, Xiaodan Liang, Renjing Pei, Guansong Lu, Songcen Xu, Wei Zhang, Hang Xu

ECCV 2024 · Image generation › Text-to-image diffusion · approach: Layout-to-image control · 15 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
add self-mask guidance (SMG) sampling (ablation)LayerDiff · MLCID test, two layersFID: 35.1 -> 21.3Improved
add self-mask guidance (SMG) sampling (ablation)LayerDiff · MLCID test, four layersFID: 83.9 -> 58.7Improved
add text-guided intra-layer attention (ablation)LayerDiff · MLCID ablation subset, two layersFID: 358.0 -> 45.8Improved
add mask dilation (kernel 5) (ablation)LayerDiff · MLCID ablation subset, two layersFID: 53.7 -> 45.8Improved
add layer-specific prompt enhancer (self+cross attention) (ablation)LayerDiff · MLCID ablation subset, three layersFID: 92.2 -> 92.9No clear effect
add layer-specific prompt enhancer (self+cross attention) (ablation)LayerDiff · MLCID ablation subset, two layersFID: 45.8 -> 46.7Hurt
replace FC mask information injection with zero-initialized scalar (ablation)LayerDiff · MLCID ablation subset, four layersFID: 140.2 -> 112.8Improved
replace FC mask information injection with zero-initialized scalar (ablation)LayerDiff · MLCID ablation subset, two layersFID: 46.7 -> 50.1Hurt
use SMG instead of SAG sampling guidance (ablation)LayerDiff · MLCID 1k prompts, all layersFID: 92.6 -> 88.9Improved
generate three and four layers with limited training dataLayerDiff · MLCID test, three and four layersFID: SD 25.9 vs LayerDiff 58.9 (three layers)Hurt

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