https://github.com/HRNet/HigherHRNet-Human-Pose-Estimation
Backbone | AP / FPS | Flip AP / FPS | Multi-scale AP / FPS |
---|---|---|---|
Hourglass-104 | 40.3 / 14 | 42.2 / 7.8 | 45.1 / 1.4 |
DLA-34 | 37.4 / 52 | 39.2 / 28 | 41.7 / 4 |
ResNet-101 | 34.6 / 45 | 36.2 / 25 | 39.3 / 4 |
ResNet-18 | 28.1 / 142 | 30.0 / 71 | 33.2 / 12 |
Backbone | AP | FPS |
---|---|---|
Hourglass-104 | 64.0 | 6.6 |
DLA-34 | 58.9 | 23 |
Backbone | FPS | AP-E | AP-M | AP-H | AOS-E | AOS-M | AOS-H | BEV-E | BEV-M | BEV-H |
---|---|---|---|---|---|---|---|---|---|---|
DLA-34 | 32 | 96.9 | 87.8 | 79.2 | 93.9 | 84.3 | 75.7 | 34.0 | 30.5 | 26.8 |
python demo.py multi_pose --demo /home/ubuntu/PycharmProjects/CenterNet-master/images/ --load_model ../models/multi_pose_dla_3x.pth
python demo.py ctdet --demo /home/ubuntu/catkin_ws/src/data_avi/navinfo_park_right.avi --load_model ../models/ctdet_coco_dla_2x.pth
python demo.py ddd --exp_id 3dop --demo /home/ubuntu/monodepth2/monodepth2-master/kitti_data/2011_09_26/2011_09_26_drive_0001_sync/image_00/data --load_model ../models/ddd_3dop.pth
https://blog.csdn.net/weixin_41765699/article/details/100118353
https://blog.csdn.net/weixin_43509263/article/details/100799415
https://www.cnblogs.com/taotingz/p/11443209.html
https://blog.csdn.net/qq_14845119/article/details/89930248
教程输入pip install requirements.txt
打开requirements.txt发现里面只有库的名字但是没有具体的版本,感觉可能是因为这个原因所以找不到要下载的版本(我猜的)
解决办法
因为requirements.txt里面的库也不多,所以每一个单独pip install 库名即可
如果感觉网速很慢的话就在末尾加上国内源,
例如:pip install 库名 -i http://pypi.douban.com/simple/ --trusted-host pypi.douban.com
参考:https://blog.csdn.net/weixin_42840933/article/details/85308265
如果依然下不了的话(例如scipy依然报错ERROR: Could not find a version that satisfies the requirement scipy (from versions: none)),那么尝试用whl的方法安装
进入到whl文件的地址,在终端输入pip install scipy-1.3.1-cp36-cp36m-manylinux1_x86_64.whl
File "XXXX/src/lib/models/networks/DCNv2/dcn_v2.py", line 13, in
import _ext as _backend
ModuleNotFoundError: No module named '_ext'kn
这是一个深度学习代码运行时报的错,错误原因是XXXX/src/lib/models/networks/DCNv2/目录下需要重新编译,要把该目录的build文件夹(如果存在的话)删除,然后在命令行运行python setup.py build develop
如果不把build文件夹删除就可能会报这样的错:
File "XXXX/src/lib/models/networks/DCNv2/dcn_v2.py", line 13, in
import _ext as _backend
ImportError: XXXX/src/lib/models/networks/DCNv2/_ext.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN2at7getTypeERKNS_6TensorE
https://blog.csdn.net/ShuqiaoS/article/details/88420326
from torch._C import *
ImportError: numpy.core.multiarray failed to import
Traceback (most recent call last):
>>> torch.__version__
'1.5.1'
pip3 install pytorch==0.4.1
/CenterNet-master/src/lib/models/networks/DCNv2$ ./make.sh
python demo.py ctdet --demo /path/to/image/or/folder/or/video --load_model ../models/ctdet_coco_dla_2x.pth
from ._dcn_v2 import lib as _lib, ffi as _ffi
ImportError: /home/ubuntu/PycharmProjects/CenterNet-master/src/lib/models/networks/DCNv2/_ext/dcn_v2/_dcn_v2.so: undefined symbol: __cudaRegisterFatBinaryEnd
sudo sh cuda_9.0.176_384.81_linux.run [更新 --0%--]---> enter
ubuntu@ubuntu-System-Product-Name:/usr/local$ vim ~/.bashrc
https://www.jianshu.com/p/6a6fbce9073f
gedit ~/.bashrc
source ~/.bashrc
vim ~/.bashrc #修改配置文件
#在文件结尾处添加
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64
export PATH=$PATH:/usr/local/cuda/bin
export CUDA_HOME=$CUDA_HOME:/usr/local/cuda
bashrc文件里设置
sudo rm -rf cuda
sudo ln -s /usr/local/cuda-9.0 /usr/local/cuda
添加软连接(symbol link)相当于添加一个快捷方式,解决cannot open shared object file: No such file or directory问题。
sudo ln -s source target
sudo ldconfig /usr/local/cuda/lib64
如果要修改软连接指向,使用-f
sudo ln -fs source target
如果是文件夹,则可以加上-n
sudo ln -fns source target
File "XXXX/src/lib/models/networks/DCNv2/dcn_v2.py", line 13, in
import _ext as _backend
ModuleNotFoundError: No module named '_ext'kn
这是一个深度学习代码运行时报的错,错误原因是XXXX/src/lib/models/networks/DCNv2/目录下需要重新编译,要把该目录的build文件夹(如果存在的话)删除,然后在命令行运行python setup.py build develop
如果不把build文件夹删除就可能会报这样的错:
File "XXXX/src/lib/models/networks/DCNv2/dcn_v2.py", line 13, in
import _ext as _backend
ImportError: XXXX/src/lib/models/networks/DCNv2/_ext.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN2at7getTypeERKNS_6TensorE
export PYTHONPATH="/usr/lib/python3.5/site-packages:$PYTHONPATH"
which python
/usr/bin/python
sudo rm /usr/bin/python
sudo ln -s /usr/bin/python3 /usr/bin/python
因为是受ROS安装的影响,ROS安装之后会在python pyth中加入:
'/opt/ros/kinetic/lib/python2.7/dist-packages'
而为了解决此问题,可以在需要运行的python文件(即使用import cv2的python文件)中,添加以下代码:
import sys
sys.path.remove('/opt/ros/kinetic/lib/python2.7/dist-packages')
通过这两行代码可以把ROS写入path中的路径给清除,进行可以import anaconda中的cv2包。
ImportError: No module named '_tkinter', please install the python3-tk package
先更新包,命令:sudo apt-get update
然后:
提示还没有'_tkinter',这时只要把python3修改成对应的python版本号,
比如我的是python3.6.5,重新输入sudo apt-get install python3.6-tk,即可
madel