Controllable Layered Image Generation for Real-World Editing

LASAGNA supports three generation modes: background-conditioned foreground generation, foreground-conditioned background generation, and text-to-all layer generation, which flexibly handle different inputs and jointly synthesize coherent, high-quality composites, backgrounds, and transparent foregrounds with realistic visual effects.

We formulate the joint generation of composite images, backgrounds, and foregrounds as a flexible, layer-conditional denoising task. This single framework supports multiple workflows, including FG_Gen, BG_Gen, and Text2All. We use a unified input representation with learnable embeddings that distinguish different roles of visual latents (noise, BG, FG, and mask) across tasks, enabling the model to adapt its behavior under various generation settings. This allows a single attention-based model to flexibly process varied combinations of inputs and targets simultaneously.

Starting with existing datasets, we implement a four-stage data construction pipeline leveraging off-the-shelf models with a custom-trained data curator. This yields a high-quality dataset as the foundation for subsequent model training.

Each row shows a composite image, a clean background, and a foreground layer with visual effects, along with corresponding captions for all components.

We compare LASAGNA with Flux.1 [1], Qwen-Image-Edit [2], and gpt-image-1[High] [3]. (a) Across three distinct generation tasks, LASAGNA consistently achieves superior inter-layer coherence and consistency. In contrast, competing models often fail to maintain these properties. (b) Moreover, by generating foregrounds with faithfully preserved visual effects, LASAGNA enables diverse post-generation editing operations on individual layers directly—a capability not supported by existing models.

We compare LASAGNA with LayerDiffuse [4]. In FG_Gen, our model produces object with appropriate size and position, along with realistic shadows consistent with the background. In BG_Gen and Text2All, our model produces visually consistent results across all layers. Furthermore, it can generate new foregrounds with corresponding visual effects, enabling flexible and realistic post-editing.

We demonstrate the benefits of explicit layer representations with visual effects by comparing three paradigms: Instruct Editing, Layer Editing, and Layer Editing with Visual Effects (LASAGNA). Across recoloring, spatial, and compositional editing tasks, the lack of explicit layer representations makes Instruct Editing prone to unintended changes and less responsive to spatial instructions, while Layer Editing with Visual Effects yields more coherent and photorealistic results.