ubuntu18.04下复现 singleshotpose（yolo-6D）源代码复现

源代码地址

首先配置好conda环境（需要提前安装ubuntu18.04和 anaconda或者miniconda）
原代码的说明里说可以选择安装python 3.6也可以装python2.7（我最开始的时候装的是python2.7版本的conda环境，但是折腾了很久，pytorch老装不上，可能是设备比较新的原因），
一直报错：

pose) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/pose/src/singleshotpose/py2$ python train.py --datacfg cfg/ape.data --modelcfg cfg/yolo-pose.cfg --initweightfile cfg/darknet19_448.conv.23 --pretrain_num_epochs 15
Traceback (most recent call last):File "train.py", line 4, in <module>import torch
ModuleNotFoundError: No module named 'torch'

我的笔记本配置是RTX2060 、i7-10875H
接下来安装python 3.6的环境

conda create -n yolo6d python=3.6

创建结果：

(base) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs$ conda create -n yolo6d python=3.6
Collecting package metadata (current_repodata.json): done
Solving environment: done## Package Plan ##environment location: /home/smile/miniconda3/envs/yolo6dadded / updated specs:- python=3.6The following packages will be downloaded:package                    |            build---------------------------|-----------------_libgcc_mutex-0.1          |      conda_forge           3 KB  conda-forge_openmp_mutex-4.5          |            1_gnu          22 KB  conda-forgeca-certificates-2021.10.8  |       ha878542_0         139 KB  conda-forgecertifi-2016.9.26          |           py36_0         217 KB  conda-forgeld_impl_linux-64-2.36.1    |       hea4e1c9_2         667 KB  conda-forgelibffi-3.4.2               |       h7f98852_5          57 KB  conda-forgelibgcc-ng-11.2.0           |      h1d223b6_11         887 KB  conda-forgelibgomp-11.2.0             |      h1d223b6_11         427 KB  conda-forgelibnsl-2.0.0               |       h7f98852_0          31 KB  conda-forgelibstdcxx-ng-11.2.0        |      he4da1e4_11         4.2 MB  conda-forgelibzlib-1.2.11             |    h36c2ea0_1013          59 KB  conda-forgencurses-6.2                |       h58526e2_4         985 KB  conda-forgeopenssl-1.1.1l             |       h7f98852_0         2.1 MB  conda-forgepip-21.3.1                 |     pyhd8ed1ab_0         1.2 MB  conda-forgepython-3.6.15              |hb7a2778_0_cpython        38.4 MB  conda-forgepython_abi-3.6             |          2_cp36m           4 KB  conda-forgereadline-8.1               |       h46c0cb4_0         295 KB  conda-forgesetuptools-49.6.0          |   py36h5fab9bb_3         936 KB  conda-forgesqlite-3.37.0              |       h9cd32fc_0         1.5 MB  conda-forgetk-8.6.11                  |       h27826a3_1         3.3 MB  conda-forgewheel-0.37.0               |     pyhd8ed1ab_1          31 KB  conda-forgexz-5.2.5                   |       h516909a_1         343 KB  conda-forgezlib-1.2.11                |    h36c2ea0_1013          86 KB  conda-forge------------------------------------------------------------Total:        55.7 MBThe following NEW packages will be INSTALLED:_libgcc_mutex      conda-forge/linux-64::_libgcc_mutex-0.1-conda_forge_openmp_mutex      conda-forge/linux-64::_openmp_mutex-4.5-1_gnuca-certificates    conda-forge/linux-64::ca-certificates-2021.10.8-ha878542_0certifi            conda-forge/linux-64::certifi-2016.9.26-py36_0ld_impl_linux-64   conda-forge/linux-64::ld_impl_linux-64-2.36.1-hea4e1c9_2libffi             conda-forge/linux-64::libffi-3.4.2-h7f98852_5libgcc-ng          conda-forge/linux-64::libgcc-ng-11.2.0-h1d223b6_11libgomp            conda-forge/linux-64::libgomp-11.2.0-h1d223b6_11libnsl             conda-forge/linux-64::libnsl-2.0.0-h7f98852_0libstdcxx-ng       conda-forge/linux-64::libstdcxx-ng-11.2.0-he4da1e4_11libzlib            conda-forge/linux-64::libzlib-1.2.11-h36c2ea0_1013ncurses            conda-forge/linux-64::ncurses-6.2-h58526e2_4openssl            conda-forge/linux-64::openssl-1.1.1l-h7f98852_0pip                conda-forge/noarch::pip-21.3.1-pyhd8ed1ab_0python             conda-forge/linux-64::python-3.6.15-hb7a2778_0_cpythonpython_abi         conda-forge/linux-64::python_abi-3.6-2_cp36mreadline           conda-forge/linux-64::readline-8.1-h46c0cb4_0setuptools         conda-forge/linux-64::setuptools-49.6.0-py36h5fab9bb_3sqlite             conda-forge/linux-64::sqlite-3.37.0-h9cd32fc_0tk                 conda-forge/linux-64::tk-8.6.11-h27826a3_1wheel              conda-forge/noarch::wheel-0.37.0-pyhd8ed1ab_1xz                 conda-forge/linux-64::xz-5.2.5-h516909a_1zlib               conda-forge/linux-64::zlib-1.2.11-h36c2ea0_1013Proceed ([y]/n)? yDownloading and Extracting Packages
libffi-3.4.2         | 57 KB     | ##################################### | 100% 
pip-21.3.1           | 1.2 MB    | ##################################### | 100% 
zlib-1.2.11          | 86 KB     | ##################################### | 100% 
_libgcc_mutex-0.1    | 3 KB      | ##################################### | 100% 
libgcc-ng-11.2.0     | 887 KB    | ##################################### | 100% 
libstdcxx-ng-11.2.0  | 4.2 MB    | ##################################### | 100% 
python-3.6.15        | 38.4 MB   | ##################################### | 100% 
_openmp_mutex-4.5    | 22 KB     | ##################################### | 100% 
readline-8.1         | 295 KB    | ##################################### | 100% 
tk-8.6.11            | 3.3 MB    | ##################################### | 100% 
ld_impl_linux-64-2.3 | 667 KB    | ##################################### | 100% 
certifi-2016.9.26    | 217 KB    | ##################################### | 100% 
setuptools-49.6.0    | 936 KB    | ##################################### | 100% 
ca-certificates-2021 | 139 KB    | ##################################### | 100% 
python_abi-3.6       | 4 KB      | ##################################### | 100% 
sqlite-3.37.0        | 1.5 MB    | ##################################### | 100% 
libnsl-2.0.0         | 31 KB     | ##################################### | 100% 
openssl-1.1.1l       | 2.1 MB    | ##################################### | 100% 
xz-5.2.5             | 343 KB    | ##################################### | 100% 
libgomp-11.2.0       | 427 KB    | ##################################### | 100% 
ncurses-6.2          | 985 KB    | ##################################### | 100% 
libzlib-1.2.11       | 59 KB     | ##################################### | 100% 
wheel-0.37.0         | 31 KB     | ##################################### | 100% 
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
#
# To activate this environment, use
#
#     $ conda activate yolo6d
#
# To deactivate an active environment, use
#
#     $ conda deactivate

然后激活conda环境：

conda activate yolo6d

然后安装pytorch 0.4.1

conda install pytorch=0.4.1 torchvision cuda90 -c pytorch

安装结果：

(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs$ conda install pytorch=0.4.1 torchvision cuda90 -c pytorch
Collecting package metadata (current_repodata.json): done
Solving environment: done## Package Plan ##environment location: /home/smile/miniconda3/envs/yolo6dadded / updated specs:- cuda90- pytorch=0.4.1- torchvisionThe following packages will be downloaded:package                    |            build---------------------------|-----------------blas-1.0                   |         openblas          46 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainca-certificates-2021.10.26 |       h06a4308_2         115 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/maincertifi-2020.6.20          |     pyhd3eb1b0_3         155 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/maincffi-1.14.6                |   py36h400218f_0         220 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainintel-openmp-2021.4.0      |    h06a4308_3561         4.2 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainjbig-2.1                   |       hdba287a_0          40 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainjpeg-9d                    |       h7f8727e_0         232 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainlibffi-3.3                 |       he6710b0_2          50 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainlibgfortran-3.0.0          |                1         281 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/freelibpng-1.6.37              |       hbc83047_0         278 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainmkl-2021.4.0               |     h06a4308_640       142.6 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainninja-1.10.2               |       h5e70eb0_2         1.5 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainnumpy-1.13.1               |     py36_nomkl_0         7.2 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/freeolefile-0.46               |           py36_0          48 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainopenblas-0.2.19            |                0         3.0 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/freepillow-4.2.1               |           py36_0         870 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/freepycparser-2.21             |     pyhd3eb1b0_0          94 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainpython-3.6.13              |       h12debd9_1        32.5 MB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/mainpytorch-0.4.1              |py36_py35_py27__9.0.176_7.1.2_2       471.7 MB  pytorchsix-1.16.0                 |     pyhd3eb1b0_0          18 KB  https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/maintorchvision-0.2.1          |             py_2          37 KB  pytorch------------------------------------------------------------Total:       665.1 MBThe following NEW packages will be INSTALLED:blas               anaconda/pkgs/main/linux-64::blas-1.0-openblascffi               anaconda/pkgs/main/linux-64::cffi-1.14.6-py36h400218f_0cuda90             pytorch/linux-64::cuda90-1.0-h6433d27_0freetype           anaconda/pkgs/free/linux-64::freetype-2.5.5-2intel-openmp       anaconda/pkgs/main/linux-64::intel-openmp-2021.4.0-h06a4308_3561jbig               anaconda/pkgs/main/linux-64::jbig-2.1-hdba287a_0jpeg               anaconda/pkgs/main/linux-64::jpeg-9d-h7f8727e_0libgfortran        anaconda/pkgs/free/linux-64::libgfortran-3.0.0-1libpng             anaconda/pkgs/main/linux-64::libpng-1.6.37-hbc83047_0libtiff            anaconda/pkgs/free/linux-64::libtiff-4.0.6-3mkl                anaconda/pkgs/main/linux-64::mkl-2021.4.0-h06a4308_640ninja              anaconda/pkgs/main/linux-64::ninja-1.10.2-h5e70eb0_2numpy              anaconda/pkgs/free/linux-64::numpy-1.13.1-py36_nomkl_0olefile            anaconda/pkgs/main/linux-64::olefile-0.46-py36_0openblas           anaconda/pkgs/free/linux-64::openblas-0.2.19-0pillow             anaconda/pkgs/free/linux-64::pillow-4.2.1-py36_0pycparser          anaconda/pkgs/main/noarch::pycparser-2.21-pyhd3eb1b0_0pytorch            pytorch/linux-64::pytorch-0.4.1-py36_py35_py27__9.0.176_7.1.2_2six                anaconda/pkgs/main/noarch::six-1.16.0-pyhd3eb1b0_0torchvision        pytorch/noarch::torchvision-0.2.1-py_2The following packages will be UPDATED:ca-certificates    anaconda/cloud/conda-forge::ca-certif~ --> anaconda/pkgs/main::ca-certificates-2021.10.26-h06a4308_2certifi            anaconda/cloud/conda-forge/linux-64::~ --> anaconda/pkgs/main/noarch::certifi-2020.6.20-pyhd3eb1b0_3The following packages will be SUPERSEDED by a higher-priority channel:libffi             anaconda/cloud/conda-forge::libffi-3.~ --> anaconda/pkgs/main::libffi-3.3-he6710b0_2python             anaconda/cloud/conda-forge::python-3.~ --> anaconda/pkgs/main::python-3.6.13-h12debd9_1Proceed ([y]/n)? yDownloading and Extracting Packages
numpy-1.13.1         | 7.2 MB    | ##################################### | 100% 
libpng-1.6.37        | 278 KB    | ##################################### | 100% 
pycparser-2.21       | 94 KB     | ##################################### | 100% 
pillow-4.2.1         | 870 KB    | ##################################### | 100% 
ninja-1.10.2         | 1.5 MB    | ##################################### | 100% 
mkl-2021.4.0         | 142.6 MB  | ##################################### | 100% 
jbig-2.1             | 40 KB     | ##################################### | 100% 
libffi-3.3           | 50 KB     | ##################################### | 100% 
certifi-2020.6.20    | 155 KB    | ##################################### | 100% 
ca-certificates-2021 | 115 KB    | ##################################### | 100% 
olefile-0.46         | 48 KB     | ##################################### | 100% 
pytorch-0.4.1        | 471.7 MB  | ##################################### | 100% 
torchvision-0.2.1    | 37 KB     | ##################################### | 100% 
six-1.16.0           | 18 KB     | ##################################### | 100% 
cffi-1.14.6          | 220 KB    | ##################################### | 100% 
intel-openmp-2021.4. | 4.2 MB    | ##################################### | 100% 
openblas-0.2.19      | 3.0 MB    | ##################################### | 100% 
libgfortran-3.0.0    | 281 KB    | ##################################### | 100% 
jpeg-9d              | 232 KB    | ##################################### | 100% 
blas-1.0             | 46 KB     | ##################################### | 100% 
python-3.6.13        | 32.5 MB   | ##################################### | 100% 
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

如果你安全上网了，可能会报错：

(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs$ conda install torchvison
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.PackagesNotFoundError: The following packages are not available from current channels:- torchvisonCurrent channels:- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/linux-64- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/noarch- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/linux-64- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/noarch- https://repo.anaconda.com/pkgs/main/linux-64- https://repo.anaconda.com/pkgs/main/noarch- https://repo.anaconda.com/pkgs/r/linux-64- https://repo.anaconda.com/pkgs/r/noarchTo search for alternate channels that may provide the conda package you're
looking for, navigate tohttps://anaconda.organd use the search bar at the top of the page.

把源码拷贝到anaconda的对应环境的文件夹下，这是我的miniconda的文件夹：/home/smile/miniconda3/envs

相关权重文件我也都放进去了：
pycharm打开源码文件夹，并且将pycharm的编译器设置为最开始我们创建的conda的编译器：

在pycharm下进行终端操作：

(base) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/yolo6d$ conda activate yolo6d

然后我按源代码的教程开始训练，发现少安装了cv2

(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/yolo6d/src$ ls
date  singleshotpose
(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/yolo6d/src$ cd singleshotpose/
(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/yolo6d/src/singleshotpose$ lsbackup         cfg.pyc       dataset.pyc              LICENSE       MeshPly.pyc                 region_loss.py    utils.pyc             valid.pybackup_multi   darknet.py    image.py                 LICENSE.txt   multi_obj_pose_estimation   region_loss.pyc   valid.ipynb           VOCdevkitcfg            darknet.pyc   image.pyc                LINEMOD       py2                         train.py         'valid _multi.ipynb'   weightfilecfg.py         dataset.py    label_file_creation.md   MeshPly.py    README.md                   utils.py         'valid _multi.py'
(yolo6d) smile@smile-X3-S-Series-GK7MR0R:~/miniconda3/envs/yolo6d/src/singleshotpose$ python train.py --datacfg cfg/ape.data --modelcfg cfg/yolo-pose.cfg --initweightfile cfg/darknet19_448.conv.23 --pretrain_num_epochs 15
Traceback (most recent call last):File "train.py", line 18, in <module>import datasetFile "/home/smile/miniconda3/envs/yolo6d/src/singleshotpose/dataset.py", line 12, in <module>from utils import read_truths_args, read_truths, get_all_filesFile "/home/smile/miniconda3/envs/yolo6d/src/singleshotpose/utils.py", line 10, in <module>import cv2
ModuleNotFoundError: No module named 'cv2'

少什么包就安装什么包：pip install opencv-python后面又报：AttributeError: module 'PIL.Image' has no attribute 'register_extensions' 到网上查了一下解决方法 ![在这里插入图片描述](https://img-blog.csdnimg.cn/450700509db443218a6f4f2099b94b02.png?x-oss-process=image/watermark,type_d3F5LXplbmhlaQ,shadow_50,text_Q1NETiBAcXFfNDQ2NDk5NDU=,size_20,color_FFFFFF,t_70,g_se,x_16) 大概意思是让安装pip install pillow、pip install image`就可以跑了
在这里插入图片描述
conda镜像源配置文件并修改
我的conda的源配置（用的清华的源）

(base) smile@smile-X3-S-Series-GK7MR0R:~$ conda config --show-sources
==> /home/smile/.condarc <==
channels:- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/- defaults
show_channel_urls: True

yolo6d 、singleshotpose数据集下载
链接: https://pan.baidu.com/s/1bD6_Dk9mlRe8WBfYSRoMeg?pwd=di5i 提取码: di5i
数据集内容如下：
在这里插入图片描述