DytanVO 代码复现(服务器端复现rtx3090)

本文主要是介绍DytanVO 代码复现(服务器端复现rtx3090)，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

源码地址

代码地址：https://github.com/castacks/DytanVO

环境配置

1.克隆github项目：

git clone https://github.com/castacks/DytanVO.git

2.利用yaml创建conda 环境：

修改yaml文件

name: dytanvo
channels:- pytorch- conda-forge
dependencies:- python=3.8- numba- tqdm- tbb- joblib- h5py- pytorch=1.7.0- torchvision=0.8.0- cudatoolkit=11.0- pip- toml=0.10.2- tomli=2.0.1- kornia=0.5.3

cd DytanVO
conda env create -f environment.yml
conda activate dytanvo

3.创建一个requirements.txt,安装相关的库

    absl-py==0.11.0antlr4-python3-runtime==4.9.3appdirs==1.4.4beautifulsoup4==4.11.1black==21.4b2cachetools==4.1.1chardet==3.0.4charset-normalizer==2.1.1cloudpickle==1.6.0cupy-cuda110cython==0.29.21data==0.4dataclasses==0.6# dcnv2==0.1decorator==5.1.1fastrlock==0.8filelock==3.8.0funcsigs==1.0.2future==0.18.2fvcore==0.1.2.post20201122gdown==4.5.1google-auth==1.23.0google-auth-oauthlib==0.4.2grpcio==1.34.0hydra-core==1.2.0idna==2.10imageio==2.9.0importlib-resources==5.9.0iopath==0.1.8joblib==0.17.0jsonpatch==1.32jsonpointer==2.3latex==0.7.0lxml==4.9.1markdown==3.3.3mypy-extensions==0.4.3# ngransac==0.0.0numpy==1.23.2oauthlib==3.1.0omegaconf==2.2.3opencv-python==4.4.0.46packaging==21.3pathspec==0.10.1portalocker==2.0.0protobuf==3.14.0pyasn1==0.4.8pyasn1-modules==0.2.8# pycocotools==2.0.4pydot==1.4.1pypng==0.0.20pysocks==1.7.1pytransform3d==1.14.0pyzmq==23.2.1regex==2022.8.17requests==2.25.0requests-oauthlib==1.3.0rsa==4.6shutilwhich==1.1.0soupsieve==2.3.2.post1splines==0.2.0tabulate==0.8.7tempdir==0.7.1tensorboard==2.4.0tensorboard-data-server==0.6.1tensorboard-plugin-wit==1.7.0timm==0.6.7toml==0.10.2torchfile==0.1.0tqdm==4.54.0trimesh==3.9.3urllib3==1.26.2visdom==0.1.8.9websocket-client==1.4.0werkzeug==1.0.1workflow==1.0zipp==3.8.1

pip install -r requirements.txt

4.编译DCNv2

cd Network/rigidmask/networks/DCNv2/; 
python setup.py install; 
cd -

下载模型和数据集

根据github的链接来下载DynaKITTI
https://drive.google.com/file/d/1BDnraRWzNf938UsfprWIkcqCSfOUyGt9/view

(另外一个数据集AirDOS-Shibuya给的链接没办法下载)

下载后解压到对应文件夹

运行

创建一个放结果的文件夹

mkdir results

创建一个run.sh的脚本，在脚本里输入（修改了一下模型名称）

traj=00_1
python -W ignore::UserWarning vo_trajectory_from_folder.py --vo-model-name vonet.pkl  \--seg-model-name segnet-kitti.pth  \--kitti --kitti-intrinsics-file data/DynaKITTI/$traj/calib.txt  \--test-dir data/DynaKITTI/$traj/image_2  \--pose-file data/DynaKITTI/$traj/pose_left.txt

运行脚本