Fast rcnn 环境配置

前言

fast rcnn是Ross B. Girshick对RCNN的一个优化,性能上有了显著提升,然而在现在看来这个模型已经相对过时,在性能较低的设备上难以做到实时.但github上的fast-rcnn项目中包含大量工具脚本,是一个caffe入门的优秀选择,在这里简单介绍一下环境的搭建以及版本问题的解决.

环境介绍

• x86_64,arm均尝试过
• ubuntu 16.04 LTS

项目下载:

• git clone –recursive https://github.com/rbgirshick/py-fast-rcnn.git
• 安装好caffe需要的所有环境
  
  • cuda
  • atlas/openBlas
  • boost-dev
  • lmdb
  • leveldb
  • protobuf,glog,gflags,hdf5
    等等,如果装不全,编译过程中会有相关错误,及时安装即可
• 在这里选择Python接口,按照说明把所有的依赖项走一遍就可以了
  for req in $(cat requirements.txt); do pip install $req; done
  
  • 然而这里的一个额外问题是fast-rcnn需要其他的库,如果缺少这些库,在运行时会报错,并且会推荐你安装这些库(形如python-xxx),这时apt install它们就可以了,大概有这些
  • python-opencv python-yaml python-gtk(??印象中有，错误由plt.show()调用报出): apt install
  • cython easydict: pip[2(可选)] install

编译

• 设置Makefile.config,这里把我使用的贴过来:

## Refer to http://caffe.berkeleyvision.org/installation.html# Contributions simplifying and improving our build system are welcome!# cuDNN acceleration switch (uncomment to build with cuDNN).#USE_CUDNN := 1# CPU-only switch (uncomment to build without GPU support).CPU_ONLY := 1# uncomment to disable IO dependencies and corresponding data layers# USE_OPENCV := 0# USE_LEVELDB := 0# USE_LMDB := 0# uncomment to allow MDB_NOLOCK when reading LMDB files (only if necessary)#   You should not set this flag if you will be reading LMDBs with any#   possibility of simultaneous read and write# ALLOW_LMDB_NOLOCK := 1# Uncomment if you're using OpenCV 3# OPENCV_VERSION := 3# To customize your choice of compiler, uncomment and set the following.# N.B. the default for Linux is g++ and the default for OSX is clang++# CUSTOM_CXX := g++# CUDA directory contains bin/ and lib/ directories that we need.CUDA_DIR := /usr/local/cuda# On Ubuntu 14.04, if cuda tools are installed via# "sudo apt-get install nvidia-cuda-toolkit" then use this instead:# CUDA_DIR := /usr# CUDA architecture setting: going with all of them.# For CUDA < 6.0, comment the *_50 lines for compatibility.CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \        -gencode arch=compute_20,code=sm_21 \        -gencode arch=compute_30,code=sm_30 \        -gencode arch=compute_35,code=sm_35 \        -gencode arch=compute_50,code=sm_50 \        -gencode arch=compute_50,code=compute_50# BLAS choice:# atlas for ATLAS (default)# mkl for MKL# open for OpenBlasBLAS := atlas# Custom (MKL/ATLAS/OpenBLAS) include and lib directories.# Leave commented to accept the defaults for your choice of BLAS# (which should work)!# BLAS_INCLUDE := /path/to/your/blas# BLAS_LIB := /path/to/your/blas# Homebrew puts openblas in a directory that is not on the standard search path# BLAS_INCLUDE := $(shell brew --prefix openblas)/include# BLAS_LIB := $(shell brew --prefix openblas)/lib# This is required only if you will compile the matlab interface.# MATLAB directory should contain the mex binary in /bin.# MATLAB_DIR := /usr/local# MATLAB_DIR := /Applications/MATLAB_R2012b.app# NOTE: this is required only if you will compile the python interface.# We need to be able to find Python.h and numpy/arrayobject.h.PYTHON_INCLUDE := /usr/include/python2.7 \        /usr/lib/python2.7/dist-packages/numpy/core/include# Anaconda Python distribution is quite popular. Include path:# Verify anaconda location, sometimes it's in root.# ANACONDA_HOME := $(HOME)/anaconda# PYTHON_INCLUDE := $(ANACONDA_HOME)/include \        # $(ANACONDA_HOME)/include/python2.7 \        # $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include \# Uncomment to use Python 3 (default is Python 2)# PYTHON_LIBRARIES := boost_python3 python3.5m# PYTHON_INCLUDE := /usr/include/python3.5m \#                 /usr/lib/python3.5/dist-packages/numpy/core/include# We need to be able to find libpythonX.X.so or .dylib.PYTHON_LIB := /usr/lib# PYTHON_LIB := $(ANACONDA_HOME)/lib# Homebrew installs numpy in a non standard path (keg only)# PYTHON_INCLUDE += $(dir $(shell python -c 'import numpy.core; print(numpy.core.__file__)'))/include# PYTHON_LIB += $(shell brew --prefix numpy)/lib# Uncomment to support layers written in Python (will link against Python libs)WITH_PYTHON_LAYER := 1# Whatever else you find you need goes here.INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serialLIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial# If Homebrew is installed at a non standard location (for example your home directory) and you use it for general dependencies# INCLUDE_DIRS += $(shell brew --prefix)/include# LIBRARY_DIRS += $(shell brew --prefix)/lib# Uncomment to use `pkg-config` to specify OpenCV library paths.# (Usually not necessary -- OpenCV libraries are normally installed in one of the above $LIBRARY_DIRS.)# USE_PKG_CONFIG := 1BUILD_DIR := buildDISTRIBUTE_DIR := distribute# Uncomment for debugging. Does not work on OSX due to https://github.com/BVLC/caffe/issues/171# DEBUG := 1# The ID of the GPU that 'make runtest' will use to run unit tests.TEST_GPUID := 0# enable pretty build (comment to see full commands)Q ?= @

• 关于cudnn的问题:由于faster-rcnn项目中使用的caffe版本过于古老,在cudnn兼容性上有一定问题,解决方法在后面会介绍
• WITH_PYTHON_LAYER这一项需要,取消注释

• 注意:INCLUDE_DIRS和LIBRARY_DIRS需要手动添加内容,为了让ld找到hdf5

• 一波操作解决编译问题

cd ${你的py-fast-rcnn}/caffe-fast-rcnnmake -jx && make pycaffe -j3

(其中x为几都行)
* 当然这里可以使用cmake,cmake配置比较繁琐,这里我们选择避开它

• 如果编译成功,会不显示任何错误信息地结束,如果没有成功显示了相关错误信息,多半是可以通过apt install解决的问题,下面介绍几个apt install解决不了的问题的解法:
  
  • ld: cannot find lcodec 这样的毛病
    将Makefile的LIBRARIES加上opencv_highgui opencv_core opencv_imgproc opencv_codec

针对cudnnv5.1的适配

由于py-fast-rcnn项目发布的时候cudnn只有4,而且不知道英伟达的工程师是如何考虑的,cudnn升级竟然连接口都随便改,这里改法也比较简单

• 下载最新的caffe： git clone https://github.com/BLVC/caffe.git

• 把${你的py-faster-rcnn目录}/caffe-fast-rcnn/src/caffe/layer/cudnn_*全部替换成最新的caffe-master中的相应目录中的相应文件

  • 替换${你的py-faster-rcnn目录}/caffe-fast-rcnn/include/caffe/util/cudnn.h

  • 最少替换如下cudnn_*文件：

    ${你的py-faster-rcnn目录}/src/caffe/layers：

    • cudnn_tanh_layer.cpp & cu
    • cudnn_sigmoid_layer.cpp & cu
    • cudnn_relu_layer.cpp & cu

    ${你的py-faster-rcnn目录}/caffe-fast-rcnn/include/caffe/layers:

    • cudnn_tanh_layer.hpp
    • cudnn_sigmoid_layer.hpp
    • cudnn_relu_layer.hpp

• 如果担心适配性问题，可以把所有含有cudnn的文件统统替换掉，包括include目录下和src目录下

训练网络

使用${你的py-faster-rcnn目录}/experiments/scripts/faster_rcnn_alt_opt.sh进行训练，参数如下

• $1: gpu序号——使用第几个显卡
• $2: 网络模型——有ZF和VGG16供选择
• $3：数据集——该脚本默认只支持pascal_voc,即VOC2007数据集

运行时问题

• TypeError: slice indices must be integers or None or have an __index__ method

  • 修改脚本，将${你的py-faster-rcnn目录}/tools/lib/minibatch.py的177行开始的循环，将其中ind,end,start变量使用int()转换为整型

• Merge(…)：AttributeError: ‘module’ object has no attribute ‘text_format’

  • 修改脚本，在train.py中靠上部位加入import google.protobuf.text_format

依旧没有解决的问题

当前安装出来的py-faster-rcnn可以测试网络，但是训练的时候会出现CURAND_CHECK失败(201 vs. 0)