File size: 1,140 Bytes
9563fbe
f97ceb0
 
 
 
 
 
 
9563fbe
 
 
 
 
b42fdda
9563fbe
 
f97ceb0
 
 
 
 
 
 
 
 
9563fbe
 
f97ceb0
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
---
license: mit
language:
  - en
pipeline_tag: depth-estimation
tags:
  - high resolution
  - monocular depth estimation
title: PatchFusion
emoji: πŸ“š
colorFrom: red
colorTo: green
sdk: gradio
sdk_version: 4.41.0
app_file: app.py
pinned: false
models:
  - zhyever/patchfusion_zoedepth
  - zhyever/patchfusion_depth_anything_vits14
  - zhyever/patchfusion_depth_anything_vitb14
  - zhyever/patchfusion_depth_anything_vitl14
  - LiheYoung/depth_anything_vits14
  - LiheYoung/depth_anything_vitb14
  - LiheYoung/depth_anything_vitl14

---

This is a demo of the high-resolution monocular depth estimation pipeline, described in the paper titled ["PatchFusion: An End-to-End Tile-Based Framework for High-Resolution Monocular Metric Depth Estimation"](https://arxiv.org/abs/2312.02284)


```
@InProceedings{li2023patchfusion,
      title={PatchFusion: An End-to-End Tile-Based Framework for High-Resolution Monocular Metric Depth Estimation},
      author={Zhenyu Li and Shariq Farooq Bhat and Peter Wonka},
      booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
      year={2024}
}
```