---
license: openrail
language:
- en
pipeline_tag: depth-estimation
tags:
- monocular depth estimation
- video depth estimation
- in-the-wild
- zero-shot
- depth
---


# 🛹 RllingDepth: Video Depth without Video Models

[![Website](doc/badges/badge-website.svg)](https://rollingdepth.github.io)
[![Hugging Face Model](https://img.shields.io/badge/GitHub-Code-blue)](https://github.com/prs-eth/RollingDepth)

This repository represents the official implementation of the paper titled "Video Depth without Video Models".

[Bingxin Ke](http://www.kebingxin.com/)<sup>1</sup>,
[Dominik Narnhofer](https://scholar.google.com/citations?user=tFx8AhkAAAAJ&hl=en)<sup>1</sup>,
[Shengyu Huang](https://shengyuh.github.io/)<sup>1</sup>,
[Lei Ke](https://www.kelei.site/)<sup>2</sup>,
[Torben Peters](https://scholar.google.com/citations?user=F2C3I9EAAAAJ&hl=de)<sup>1</sup>,
[Katerina Fragkiadaki](https://www.cs.cmu.edu/~katef/)<sup>2</sup>,
[Anton Obukhov](https://www.obukhov.ai/)<sup>1</sup>,
[Konrad Schindler](https://scholar.google.com/citations?user=FZuNgqIAAAAJ&hl=en)<sup>1</sup>


<sup>1</sup>ETH Zurich, 
<sup>2</sup>Carnegie Mellon University



<!-- internal checkpoint id: 54-1_tar+hyp_1-3frame_d-pertub_d-squeeze_bs8/checkpoint/iter_022000 -->
These components are copied from [Stable-Diffusion-2](https://huggingface.co/stabilityai/stable-diffusion-2): `text_encoder`, `tokenizer`, and `vae`.


<!-- ## 🎓 Citation

```bibtex

``` -->

## 🎫 License


This code of this work is licensed under the Apache License, Version 2.0 (as defined in the [LICENSE-CODE](LICENSE-CODE.txt)).

The model is licensed under RAIL++-M License (as defined in the [LICENSE-MODEL](LICENSE-MODEL.txt))

By downloading and using the code and model you agree to the terms in [LICENSE-CODE](LICENSE-CODE.txt) and [LICENSE-MODEL](LICENSE-MODEL.txt) respectively.