Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models

Abstract

We present the first framework to solve linear inverse problems leveraging pre-trained \textit{latent} diffusion models. Previously proposed algorithms (such as DPS and DDRM) only apply to \textit{pixel-space} diffusion models. We theoretically analyze our algorithm showing provable sample recovery in a linear model setting. The algorithmic insight obtained from our analysis extends to more general settings often considered in practice. Experimentally, we outperform previously proposed posterior sampling algorithms in a wide variety of problems including random inpainting, block inpainting, denoising, deblurring, destriping, and super-resolution.

Cite

Text

Rout et al. "Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models." Neural Information Processing Systems, 2023.

Markdown

[Rout et al. "Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models." Neural Information Processing Systems, 2023.](https://mlanthology.org/neurips/2023/rout2023neurips-solving/)

BibTeX

@inproceedings{rout2023neurips-solving,
  title     = {{Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models}},
  author    = {Rout, Litu and Raoof, Negin and Daras, Giannis and Caramanis, Constantine and Dimakis, Alex and Shakkottai, Sanjay},
  booktitle = {Neural Information Processing Systems},
  year      = {2023},
  url       = {https://mlanthology.org/neurips/2023/rout2023neurips-solving/}
}