<div align="center">
<h2>AFNet
: Adaptive Fusion of Single-View and Multi-View Depth for Au
tonomous Driving</h2>
<h2>**CVPR 2024**<h2>
[Paper]
(https
://arxiv.org/pdf/2403.07535.pdf
)
</div>
This work presents AFNet, a new multi-view and singleview depth fusion network AFNet for alleviating
the defects of
the existing multi-view methods, which will fail under noisy poses in real-world au
tonomous driving scenarios.
![teaser]
(assets/pointcloud2.png
)
## ✏️ Changelog
### Mar. 20 2024
* Initial release. Due
to the confidentiality agreement,
the accuracy of
the current reproduced model on KITTI is very slightly different from that in
the paper. We release an initial version first, and
the final version will be released soon.
* In addition,
the models trained under noise pose will soon be released.
## ⚙️ Installation
The code is tested with CUDA11.7. Please use
the following commands
to install dependencies
:
```
conda create --name AFNet
python=3.7
conda activate AFNet
pip install -r requirements.txt
```
## 🎬 Demo
![teaser]
(assets/visual_compare.png
)
## ⏳ Training & Testing
We use 4 Nvidia 3090 GPU for training. You may need
to modify 'CUDA_VISIBLE_DEVICES' and batch size
to accommodate your GPU resources.
#### Training
First download and
extract DDAD and KITTI data and split. You should download and process DDAD dataset follow [DDAD🔗]
(https
://github.com/TRI-ML/DDAD
).
#### Download
[__split__ 🔗]
(https
://1drv.ms/u/s!AtFfCZ2Ckf3DghYrBvQ-DWCQR1Nd?e=Q6qz8d
) （You need
to move this json
file in split
to the data_split path）
[ models 🔗]
(https
://1drv.ms/u/s!AtFfCZ2Ckf3DghVXXZY611mqxa8B?e=nZ7taR
) （models for testing）
Then run
the following command
to train our model.
```
bash scripts/train.sh
```
#### Testing
First download and
extract data, split and pretrained models.
### DDAD
:
run
:
```
python eval_ddad.py --cfg "./configs/DDAD.conf"
```
You should get something like
these
:
| abs_rel | sq_rel | log10 | rmse | rmse_log | a1 | a2 | a3 | abs_diff |
|---------|--------|-------|-------|----------|-------|-------|-------|----------|
| 0.088 | 0.979 | 0.035 | 4.60 | 0.154 | 0.917 | 0.972 | 0.987 | 2.042 |
### KITTI
:
run
:
```
python eval_kitti.py --cfg "./configs/kitti.conf"
```
You should get something like
these
:
| abs_rel | sq_rel | log10 | rmse | rmse_log | a1 | a2 | a3 | abs_diff |
|---------|--------|-------|-------|----------|-------|-------|-------|----------|
| 0.044 | 0.132 | 0.019 | 1.712 | 0.069 | 0.980 | 0.997 | 0.999 | 0.804 |
#### Acknowledgement
Thanks
to Zhenpei Yang for opening source of his excellent works [MVS2D]
(https
://github.com/zhenpeiyang/MVS2D?tab=readme-ov-
file#nov-27-2021
)
## Citation
If you find this project useful, please consider citing
:
```bibtex
@misc{cheng2024adaptive,
title={Adaptive Fusion of Single-View and Multi-View Depth for Au
tonomous Driving},
author={JunDa Cheng and Wei Yin and Kaixuan Wang and Xiaozhi Chen and Shijie Wang and Xin Yang},
year={2024},
eprint={2403.07535},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
PS E
:\AFNet-main> pip install -r requirements.txt
Defaulting
to user installation because normal site-packages is not writeable
Collecting imageio==2.9.0
(from -r requirements.txt
(line 1
))
Downloading imageio-2.9.0-py3-none-any.whl.metadata
(2.6 kB
)
Collecting ipdb==0.12.3
(from -r requirements.txt
(line 2
))
Downloading ipdb-0.12.3.tar.gz
(10 kB
)
Preparing metadata
(setup.py
) ... done
Collecting joblib==1.0.1
(from -r requirements.txt
(line 3
))
Downloading joblib-1.0.1-py3-none-any.whl.metadata
(6.8 kB
)
Collecting lz4==3.1.3
(from -r requirements.txt
(line 4
))
Downloading lz4-3.1.3.tar.gz
(159 kB
)
Preparing metadata
(setup.py
) ... done
Collecting matplotlib==3.1.3
(from -r requirements.txt
(line 5
))
Downloading matplotlib-3.1.3.tar.gz
(40.9 MB
)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 40.9/40.9 MB 6.2 MB/s eta 0
:00
:00
Preparing metadata
(setup.py
) ...
error
error: subprocess-exited-with-
error
×
python setup.py
egg_info did not run successfully.
│ exit code
: 1
╰─> [28 lines of output]
C
:\Users\HP\AppData\Local\Temp\pip-install-xl7u4dpy\matplotlib_e71ff03b933d48d99aa6e0525941cb19\setup.py
:32
: Setup
toolsDeprecationWarning
: The test command is disabled and references
to it are deprecated.
!!
********************************************************************************
Please remove any references
to `setup
tools.command.test` in all supported versions of
the affected package.
This deprecation is overdue, please update your project and remove deprecated
calls
to avoid build
errors in
the future.
********************************************************************************
!!
from setup
tools.command.test import test as TestCommand
Traceback
(most recent call last
):
File "<string>", line 2, in <module>
File "<pip-setup
tools-caller>", line 34, in <module>
File "C
:\Users\HP\AppData\Local\Temp\pip-install-xl7u4dpy\matplotlib_e71ff03b933d48d99aa6e0525941cb19\setup.py", line 52, in <module>
__version__ = versioneer.get_version
()
^^^^^^^^^^^^^^^^^^^^^^^^
File "C
:\Users\HP\AppData\Local\Temp\pip-install-xl7u4dpy\matplotlib_e71ff03b933d48d99aa6e0525941cb19\versioneer.py", line 1410, in get_version
return get_versions
()["version"]
^^^^^^^^^^^^^^
File "C
:\Users\HP\AppData\Local\Temp\pip-install-xl7u4dpy\matplotlib_e71ff03b933d48d99aa6e0525941cb19\versioneer.py", line 1344, in get_versions
cfg = get_config_from_root
(root
)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C
:\Users\HP\AppData\Local\Temp\pip-install-xl7u4dpy\matplotlib_e71ff03b933d48d99aa6e0525941cb19\versioneer.py", line 401, in get_config_from_root
parser = configparser.SafeConfigParser
()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Attribute
Error: module 'configparser' has no attribute 'SafeConfigParser'. Did you mean
: 'RawConfigParser'?
[end of output]
note
: This
error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed
× Encountered
error while generating package metadata.
╰─> See above for output.
note
: This is an issue with
the package mentioned above, not pip.
hint
: See above for details.
解决Python Egg Cache权限错误
本文详细介绍了如何解决Python在尝试提取文件到egg缓存目录时遇到的权限问题，包括设置PYTHON_EGG_CACHE环境变量来指向可访问的目录。
扫一扫
1197
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
