AI Paper Atlas

Research map / Image generation / Editing and style transfer

Qwen-Image-Layered: Towards Inherent Editability via Layer Decomposition

Shengming Yin, Zekai Zhang, Zecheng Tang, Kaiyuan Gao, Xiao Xu, Kun Yan, Jiahao Li, Yilei Chen, Yuxiang Chen, Heung-Yeung Shum, Lionel M. Ni, Junyang Lin and 1 more

CVPR 2026 · Image generation › Editing and style transfer · approach: Instruction-based diffusion editing · 39 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 Layer3D RoPE (vs standard 2D RoPE) (ablation)Qwen-Image-Layered-I2L · Crello (max layer merge 0)Alpha soft IoU: 0.3725 -> 0.5844Improved
add RGBA-VAE encoding for the input image (vs RGB VAE) (ablation)Qwen-Image-Layered-I2L · Crello (max layer merge 0)Alpha soft IoU: 0.5844 -> 0.6504Improved
add RGBA-VAE encoding for the input image (vs RGB VAE) (ablation)Qwen-Image-Layered-I2L · Crello (max layer merge 0)RGB L1: 0.1894 -> 0.1649Improved
add multi-stage training (vs init directly from text-to-image weights) (ablation)Qwen-Image-Layered-I2L · Crello (max layer merge 0)Alpha soft IoU: 0.6504 -> 0.8705Improved
add multi-stage training (vs init directly from text-to-image weights) (ablation)Qwen-Image-Layered-I2L · Crello (max layer merge 0)RGB L1: 0.1649 -> 0.0594Improved
use end-to-end Qwen-Image-Layered-I2L instead of LayerDimage decomposition · Crello (max layer merge 0)Alpha soft IoU: 0.7520 -> 0.8705Improved
use end-to-end Qwen-Image-Layered-I2L instead of LayerDimage decomposition · Crello (max layer merge 5)RGB L1: 0.0396 -> 0.0363Improved
use RGBA-VAE instead of AlphaVAE (FLUX)RGBA reconstruction · AIM-500PSNR: 36.9439 -> 38.8252Improved
use RGBA-VAE instead of AlphaVAE (FLUX)RGBA reconstruction · AIM-500rFID: 11.7884 -> 5.3132Improved
fine-tune on Crello training set before evaluationQwen-Image-Layered-I2L · CrelloAlpha soft IoUNo 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