BiFuse - GitLite

KO | EN

gitlite — search

#javascript #ai #android #rust #deep-learning #docker #java #windows #awesome #awesome-list #vue #security #mcp #openai

BiFuse

(★ 181)

[CVPR2020] BiFuse: Monocular 360 Depth Estimation via Bi-Projection Fusion

File Explorer

Download Latest Version (.zip)

- __init__.py
- FCRN.py
- resnet.py
- Sample.jpg
- 007.jpg
- 146.jpg
- 147.jpg
- 1690-teaser.gif
- 200.jpg
- 232.jpg
- 236.jpg
- 246.jpg
- 260.jpg
- 272.jpg
- 279.jpg
- fig.png
- README.md
- utils.py
- vis3D.py
- - __init__.py
  - Cube2Equirec.py
- - .gitignore
  - __init__.py
  - Equirec2Cube.py
  - EquirecRotate.py
  - EquirecRotate2.py
- __init__.py
- CETransform.py
- CubePad.py
- ModelSaver.py
- SpherePad.py
- Transform.py
.gitignore
LICENSE
main.py
README.md
requirements.txt

# Installation Guide

1. Get the code

git clone https://github.com/yuhsuanyeh/BiFuse

Downloads the entire project code from GitHub to your computer.

cd BiFuse

Moves into the project folder you just downloaded.

2. Python

Easy Recommended

Prerequisites

Git Needed to download the project code from GitHub.
Python 3 On Windows, be sure to check 'Add Python to PATH' during installation.

pip install -r requirements.txt

Installs the Python libraries listed in requirements.txt (or similar).

✅ If it runs without errors and prints output in the terminal, it worked.

Pulled directly from this repo's README.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# [CVPR2020] BiFuse: Monocular 360 Depth Estimation via Bi-Projection Fusion

<p align='center'>
<img src='src/1690-teaser.gif'>
</p>

### [[Paper](http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_BiFuse_Monocular_360_Depth_Estimation_via_Bi-Projection_Fusion_CVPR_2020_paper.pdf)] [[Project Page](https://fuenwang.ml/project/bifuse/)]

## Getting Started
#### Requirements
- Python (tested on 3.7.4)
- PyTorch (tested on 1.4.0)
- Other dependencies
```
pip install -r requirements.txt
```

## Usage
First clone our repo:
```
git clone https://github.com/Yeh-yu-hsuan/BiFuse.git
cd BiFuse
```
### Step1
Download our [pretrained Model](https://drive.google.com/file/d/1EOEfyVuaJC1k5xAtqG37yXHxN-LnxA2n/view?usp=sharing) and create a **save** folder:
```
mkdir save
```
then put the ```BiFuse_Pretrained.pkl``` into **save** folder.
### Step2
**My_Test_Data** folder has contained a ```Sample.jpg``` RGB image as an example. <br> 
If you want to test your own data, please put your own rgb images into **My_Test_Data** folder and run:
```
python main.py --path './My_Test_Data'
```
Our argument: <br>
**```--path```** is the folder path of your own testing images.  <br>
**```--nocrop```**  if you don't want to crop the original images. <br>

After testing, you can see the results in **My_Test_Result** folder! <br>
+ Here shows some sample results
<p float="left">
  <img src="src/007.jpg" width="30%" align="middle"/>
  <img src="src/146.jpg" width="30%" align="middle"/>
  <img src="src/147.jpg" width="30%" align="middle"/>
</p>

<p float="left">
  <img src="src/200.jpg" width="30%" align="middle"/>
  <img src="src/232.jpg" width="30%" align="middle"/>
  <img src="src/246.jpg" width="30%" align="middle"/>
</p>

<p float="left">
  <img src="src/260.jpg" width="30%" align="middle" />
  <img src="src/272.jpg" width="30%" align="middle" />
  <img src="src/236.jpg" width="30%" align="middle" />
</p>

The Restuls contain ```Combine.jpg```, ```Depth.jpg```, and ```Data.npy```. <br>
```Combine.jpg``` is concatenating rgb image with its corresponding depth map prediction. <br>
```Depth.jpg``` is only depth map prediction. <br>
```Data.npy``` is the original data of both RGB and predicted depth value. <br>

### Point Cloud Visualization
If you also want to visualize the point cloud of predicted depth, we also provide the script to render it. You can have a look at [tools/](https://github.com/Yeh-yu-hsuan/BiFuse/tree/master/tools).

## License
This work is licensed under MIT License. See [LICENSE](LICENSE) for details.

If you find our code/models useful, please consider citing our paper:
```
@InProceedings{BiFuse20,
author = {Wang, Fu-En and Yeh, Yu-Hsuan and Sun, Min and Chiu, Wei-Chen and Tsai, Yi-Hsuan},
title = {BiFuse: Monocular 360 Depth Estimation via Bi-Projection Fusion},
booktitle = {The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}
```