English

SVGFusion: A VAE-Diffusion Transformer for Vector Graphic Generation

Computer Vision and Pattern Recognition 2026-04-10 v3 Graphics Machine Learning

Abstract

Generating high-quality Scalable Vector Graphics (SVGs) from text remains a significant challenge. Existing LLM-based models that generate SVG code as a flat token sequence struggle with poor structural understanding and error accumulation, while optimization-based methods are slow and yield uneditable outputs. To address these limitations, we introduce SVGFusion, a unified framework that adapts the VAE-diffusion architecture to bridge the dual code-visual nature of SVGs. Our model features two core components: a Vector-Pixel Fusion Variational Autoencoder (VP-VAE) that learns a perceptually rich latent space by jointly encoding SVG code and its rendered image, and a Vector Space Diffusion Transformer (VS-DiT) that achieves globally coherent compositions through iterative refinement. Furthermore, this architecture is enhanced by a Rendering Sequence Modeling strategy, which ensures accurate object layering and occlusion. Evaluated on our novel SVGX-Dataset comprising 240k human-designed SVGs, SVGFusion establishes a new state-of-the-art, generating high-quality, editable SVGs that are strictly semantically aligned with the input text.

Keywords

Cite

@article{arxiv.2412.10437,
  title  = {SVGFusion: A VAE-Diffusion Transformer for Vector Graphic Generation},
  author = {Ximing Xing and Juncheng Hu and Ziteng Xue and Jing Zhang and Buyu Li and Sheng Wang and Dong Xu and Qian Yu},
  journal= {arXiv preprint arXiv:2412.10437},
  year   = {2026}
}

Comments

project page: https://ximinng.github.io/SVGFusionProject/

R2 v1 2026-06-28T20:34:36.980Z