Markup-to-Image Diffusion Models with Scheduled Sampling

Method and contribution

Learns to render LaTeX, HTML table layouts, LilyPond music notation, and SMILES molecules with conditional diffusion. Scheduled sampling exposes denoisers to their own generated inputs to address compounding errors.

The original diffusion-based markup-to-image method and four-domain rendering benchmark discussed in this guide.

Abstract

Building on recent advances in image generation, we present a fully data-driven approach to rendering markup into images. The approach is based on diffusion models, which parameterize the distribution of data using a sequence of denoising operations on top of a Gaussian noise distribution. We view the diffusion denoising process as a sequential decision making process, and show that it exhibits compounding errors similar to exposure bias issues in imitation learning problems. To mitigate these issues, we adapt the scheduled sampling algorithm to diffusion training. We conduct experiments on four markup datasets: mathematical formulas (LaTeX), table layouts (HTML), sheet music (LilyPond), and molecular images (SMILES). These experiments each verify the effectiveness of the diffusion process and the use of scheduled sampling to fix generation issues. These results also show that the markup-to-image task presents a useful controlled compositional setting for diagnosing and analyzing generative image models.

Read in context

Learning to render LaTeX, SMILES, HTML and music notation →

Citation

Yuntian Deng, Noriyuki Kojima, Alexander M. Rush. Markup-to-Image Diffusion Models with Scheduled Sampling. ICLR 2023.

@inproceedings{deng2023markuptoimage,
  title = {{Markup-to-Image Diffusion Models with Scheduled Sampling}},
  author = {Deng, Yuntian and Kojima, Noriyuki and Rush, Alexander M.},
  year = {2023},
  booktitle = {ICLR 2023},
  url = {https://arxiv.org/abs/2210.05147},
  eprint = {2210.05147},
  archivePrefix = {arXiv},
  doi = {10.48550/arXiv.2210.05147}
}