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

TL;DR

Qwen-Image-Layered introduces an end-to-end diffusion model that decomposes images into editable, semantically disentangled RGBA layers, enabling consistent and isolated image editing akin to professional design tools.

Contribution

It proposes a novel multilayer image decomposition framework with variable layer handling, supported by a new pipeline for extracting multilayer images from Photoshop files.

Findings

Outperforms existing methods in decomposition quality

Enables inherently editable image representations

Establishes a new paradigm for consistent image editing

Abstract

Recent visual generative models often struggle with consistency during image editing due to the entangled nature of raster images, where all visual content is fused into a single canvas. In contrast, professional design tools employ layered representations, allowing isolated edits while preserving consistency. Motivated by this, we propose \textbf{Qwen-Image-Layered}, an end-to-end diffusion model that decomposes a single RGB image into multiple semantically disentangled RGBA layers, enabling \textbf{inherent editability}, where each RGBA layer can be independently manipulated without affecting other content. To support variable-length decomposition, we introduce three key components: (1) an RGBA-VAE to unify the latent representations of RGB and RGBA images; (2) a VLD-MMDiT (Variable Layers Decomposition MMDiT) architecture capable of decomposing a variable number of image layers; and (3) a Multi-stage Training strategy to adapt a pretrained image generation model into a multilayer image decomposer. Furthermore, to address the scarcity of high-quality multilayer training images, we build a pipeline to extract and annotate multilayer images from Photoshop documents (PSD). Experiments demonstrate that our method significantly surpasses existing approaches in decomposition quality and establishes a new paradigm for consistent image editing. Our code and models are released on \href{https://github.com/QwenLM/Qwen-Image-Layered}{https://github.com/QwenLM/Qwen-Image-Layered}