table of Contents
Foreword
In this paper, WSL (windows subsystem for linux), mounted on Anconda3 Caffe-faster-rcnn CPU version. Plan is built on the notebook CPU environment for understanding the code, and reasoning tests. Model training program implemented by the cloud platform.
By creating a virtual environment conda
conda create -n caffe_py27 python=2.7
Tip: update the relevant source system, conda source
subsequent operation enters caffe_py27
conda activate caffe_py27
Git repository to download Faster rcnn
Original download instructions: git clone --recursive https://github.com/rbgirshick/py-faster-rcnn.git
since github slow download, download code using a cloud (https://gitee.com/).
git clone https://gitee.com/shaozhechen/py-faster-rcnn.git
cd ./py-faster-rcnn/
Download the FAST-rcnn-Caffe ( ! Note that the author library @ 0dcd397 for the corresponding uploaded version! I repeat installed several times only to find ... )
git clone https://gitee.com/liu1guo2qiang3/caffe-fast-rcnn.git
cd ./caffe-fast-rcnn
git checkout 0dcd397 #这个同步到指定版本的commit
Install python-dependent
Enter. \ Py-faster-rcnn \ caffe-fast-rcnn \ python directory
pip install -r requirements.txt
提示：
You are using pip version 9.0.1, however version 20.0.2 is available.
You should consider upgrading via the ‘pip install --upgrade pip’ command.
pip install --upgrade pip
PIP Collecting
Downloading https://files.pythonhosted.org/packages/54/0c/d01aa759fdc501a58f431eb594a17495f15b88da142ce14b5845662c13f3/pip-20.0.2-py2.py3-none-any.whl (1.4MB)
4% | █▍ | 61kB 4.2 kb / s eta 0: 05: 33E
network link interruption
can be directly copied by Thunder download link, using the following command to install (installation files need to be copied to the current directory)
pip install pip-20.0.2-py2.py3-none-any.whl
Many are due to network timeouts lead to bad; you can try increasing the latency
sudo pip install --default-timeout=100 future
Tips
from the command update
pip install --upgarde python-dateutil
The update is successful
Additionally, you need to modify the relevant version of requirements.txt: python-dateutil> = 1.4, <2.9
has many libraries online update is too slow, basically by Thunder download, and then install; part csdn uploaded to the resource, to download.
Compile configuration changes
Enter the directory. \ Py-faster-rcnn \ caffe-fast-rcnn
copy the configuration file
cp Makefile.config.example Makefile.config
Remove comments
CPU_ONLY := 1
WITH_PYTHON_LAYER := 1
Path Configuration python (based Anaconda3 python2.7)
ANACONDA_HOME := $(HOME)/anaconda3/envs/caffe_py27
PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
$(ANACONDA_HOME)/include/python2.7 \
$(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include \
PYTHON_LIB := $(ANACONDA_HOME)/lib
Add hd5f library path
/usr/include/hdf5/serial及/usr/lib/x86_64-linux-gnu/hdf5/serial
INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial
ps: I do not know python path:
python #进入python脚本
import sys
print(sys.executable)
Compiled lib
Enter the directory \ py-faster-rcnn \ lib
modify setup.py content
annotation corresponding to the following
#CUDA = locate_cuda() only cpu
#self.set_executable('compiler_so', CUDA['nvcc']) only cpu
# Extension('nms.gpu_nms',
# ['nms/nms_kernel.cu', 'nms/gpu_nms.pyx'],
# library_dirs=[CUDA['lib64']],
# libraries=['cudart'],
# language='c++',
# runtime_library_dirs=[CUDA['lib64']],
# this syntax is specific to this build system
# we're only going to use certain compiler args with nvcc and not with
# gcc the implementation of this trick is in customize_compiler() below
# extra_compile_args={'gcc': ["-Wno-unused-function"],
# 'nvcc': ['-arch=sm_35',
# '--ptxas-options=-v',
# '-c',
# '--compiler-options',
# "'-fPIC'"]},
# include_dirs = [numpy_include, CUDA['include']]
# ),
Then make
prompt:
pip install easydict
提示：
ImportError: No module named cv2
pip install opencv-python
编译Caffe
进入目录…\py-faster-rcnn\caffe-fast-rcnn
make
make pycaffe
正常不会报错，如果报错查看是否缺少东西或者版本冲突
测试
需要下载模型：
./data/scripts/fetch_faster_rcnn_models.sh
外网下载过慢：
提供百度盘下载链接：
链接: https://pan.baidu.com/s/13yZmsgZI72sAwyKbBzzKEw 提取码: jmz6
由于使用cpu，需要修改lib中的相关内容
/lib/fast_rcnn/nms_wrapper.py
修改结果如下：
from fast_rcnn.config import cfg
#from nms.gpu_nms import gpu_nms
from nms.cpu_nms import cpu_nms
def nms(dets, thresh, force_cpu=False):
"""Dispatch to either CPU or GPU NMS implementations."""
if dets.shape[0] == 0:
return []
# if cfg.USE_GPU_NMS and not force_cpu:
# return gpu_nms(dets, thresh, device_id=cfg.GPU_ID)
else:
return cpu_nms(dets, thresh)
进入.\py-faster-rcnn
python .tools/demo.py --cpu
最终成功！！！
Loaded network /mnt/d/9_deeplearning/caffe_rcnn_try/py-faster-rcnn/data/faster_rcnn_models/VGG16_faster_rcnn_final.caffemodel
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Demo for data/demo/000456.jpg
Detection took 29.743s for 300 object proposals
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Demo for data/demo/000542.jpg
Detection took 21.961s for 161 object proposals
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Demo for data/demo/001150.jpg
Detection took 22.609s for 194 object proposals
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Demo for data/demo/001763.jpg
Detection took 20.895s for 196 object proposals
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Demo for data/demo/004545.jpg
Detection took 22.609s for 300 object proposals
PS：
下图出现关于“roi_pooling_param”是caffe版本错误，需要用作者对应上传的“ 0版本dcd397”
关于多个protobuf库
可以通过protoc --version 查看版本；
此外还可以查看安装了有哪些版本
conda list 和pip list
上述两个命令可能出现不同的版本，可以卸载不需要的版本
conda 中protobuf库包括:libprotobuf以及protobuf
conda remove libprotobuf
pip uninstall protobuf
I installed the final version 2.6.1 protobuf
by pip install protobuf == 2.6.1 installed successfully
Reference links:
[1]. Https://www.jianshu.com/p/30fcc3df764d
