cs.AINov 19, 2026

ProtoDiffusion: Token-per-Layer Generative Modeling of Multi-Layer Images

Authors: Adrià Ciurana Lanau, Harleen Hanspal, Lusine Abrahamyan, Emilio Tylson Baixauli, Yu Cao, Danai Triantafyllidou, Stefano Zorzi, Gustavo Zanoni Felipe, 

Organizations: Canva · Queen Mary University of London · Universidade Estadual de Maring´a

Abstract

Graphic designs are ordered compositions of RGBA layers whose separation enables selection, movement, and reuse. Existing layered models represent each layer with a full-canvas or region-sized spatial latent grid, so representation size grows with the cumulative spatial extent of the layers. We present ProtoDiffusion, a latent-diffusion framework built on ProtoVAE, a multi-layer variational autoencoder that encodes a design into one shared spatial prototype and one candidate vector per layer. Each candidate conditions a shared decoder to reconstruct its corresponding RGBA layer. For NN layers, the representation contains HW+NHW+N tokens before backbone-specific packing, compared with NHWNHW for full-grid-per-layer representations. ProtoDiffusion jointly denoises the prototype and a fixed set of candidate slots, conditioned on an image or text, with bounding boxes as optional layout guidance. The model leaves unneeded slots empty and determines the number of layers and, in absence of boxes, their layout, requiring neither as input. The same architecture supports both image-to-design and text-to-design. ProtoVAE achieves high-fidelity reconstruction from its compact shared latent representation, reaching 40.43 dB PSNR on layer crops. On the LayerD evaluation split, image-to-design achieves RGB L1 scores of 0.040 with bounding boxes and 0.057 without, with box-guided alpha soft IoU (0.850) comparable to the strongest fine-tuned baseline (0.871). On merged images, text-to-design reaches 66.41 FID versus 86.84 for FLUX.2 [klein], achieves a 0.860 text-accuracy score, and runs 19.1×19.1\times faster than the same-backbone generation-then-decomposition cascade.

Figures & tables

Appendix figures & tables11 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. LaDe: Unified Multi-Layered Graphic Media Generation and Decomposition

    Mar 18, 2026Vlad-Constantin Lungu-Stan, Ionut Mironica, Mariana-Iuliana GeorgescuGraphic Design GenerationImage Generation

  2. PixelART: Image-to-Layer Decomposition without Latents or Text-to-Image Pretraining

    Sep 19, 2026Zelin Jia, Zhao Zhang, Zhicong Tang +2Pixel-Space Diffusion ModelsDiffusion Transformer