README.md

metadata

license: cc-by-sa-4.0
language:
  - en
pipeline_tag: depth-estimation
tags:
  - monocular
  - depth estimation
  - single image depth estimation
  - single image
  - in-the-wild
  - zero-shot
  - depth

Marigold: Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation

This model represents the official checkpoint of the paper titled "Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation".

Bingxin Ke, Anton Obukhov, Shengyu Huang, Nando Metzger, Rodrigo Caye Daudt, Konrad Schindler

We present Marigold, a diffusion model and associated fine-tuning protocol for monocular depth estimation. Its core principle is to leverage the rich visual knowledge stored in modern generative image models. Our model, derived from Stable Diffusion and fine-tuned with synthetic data, can zero-shot transfer to unseen data, offering state-of-the-art monocular depth estimation results.

🎓 Citation

@misc{ke2023marigold,
  author    = {Ke, Bingxin and Obukhov, Anton and Huang, Shengyu and Metzger, Nando and Daudt, Rodrigo Caye and Schindler, Konrad},
  title     = {Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation},
  year      = {2023},
}

License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.