Autoware 笔记No.8 ENet 障碍物识别（vision segment ENet detect）

一. 前言

人人为我，我为人人

我不喜欢讲一些网络上抄的网络模型，全是干货，让大家直接上手干，如果想探讨可以联系我。

请大家按照我的Autoware 1.14安装

二. 安装

（1）下载ENet，一定要安装在home目录下，否则vision_segment_enet_detect.launch文件中的network_definition_file和pretrained_model_file路径会有变化。

$ cd ~
$ git clone --recursive https://github.com/TimoSaemann/ENet.git
$ cd ENet/caffe-enet

（2）修改Makefile.config，我用的CUDA10.0，直接贴出我的Makefile.config

## 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 := 1
USE_LEVELDB := 1
USE_LMDB := 1

# 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-10.0
# 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.
# For CUDA < 8.0, comment the *_60 and *_61 lines for compatibility.
# For CUDA >= 9.0, comment the *_20 and *_21 lines for compatibility.
CUDA_ARCH := -gencode arch=compute_61,code=compute_61

# BLAS choice:
# atlas for ATLAS (default)
# mkl for MKL
# open for OpenBlas
BLAS := open
# Custom (MKL/ATLAS/OpenBLAS) include and lib directories.
# Leave commented to accept the defaults for your choice of BLAS
# (which should work)!
BLAS_INCLUDE := /usr/local/cuda-10.0/targets/x86_64-linux/include
BLAS_LIB := /usr/local/cuda-10.0/targets/x86_64-linux/lib

# 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/serial/
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu

# 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 := 1

# N.B. both build and distribute dirs are cleared on `make clean`
BUILD_DIR := build
DISTRIBUTE_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 ?= @

（3）修改Makefile

打开Makefile，查找LIBRARIES，修改为：

LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial

（4）编译ENet

make && make distribute

如果出现错误如下：

./include/caffe/util/cudnn.hpp:113:70: error: too few arguments to function ‘cudnnStatus_t cudnnSetConvolution2dDescriptor(cudnnConvolutionDescriptor_t, int, int, int, int, int, int, cudnnConvolutionMode_t, cudnnDataType_t)’
       pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION));

修改/home/xxx/ENet/caffe-enet/include/caffe/util/cudnn.hpp为如下代码。

#ifndef CAFFE_UTIL_CUDNN_H_
#define CAFFE_UTIL_CUDNN_H_
#ifdef USE_CUDNN

#include <cudnn.h>

#include "caffe/common.hpp"
#include "caffe/proto/caffe.pb.h"

#define CUDNN_VERSION_MIN(major, minor, patch) 
    (CUDNN_VERSION >= (major * 1000 + minor * 100 + patch))

#define CUDNN_CHECK(condition) 
  do { 
    cudnnStatus_t status = condition; 
    CHECK_EQ(status, CUDNN_STATUS_SUCCESS) << " "
      << cudnnGetErrorString(status); 
  } while (0)

inline const char* cudnnGetErrorString(cudnnStatus_t status) {
  switch (status) {
    case CUDNN_STATUS_SUCCESS:
      return "CUDNN_STATUS_SUCCESS";
    case CUDNN_STATUS_NOT_INITIALIZED:
      return "CUDNN_STATUS_NOT_INITIALIZED";
    case CUDNN_STATUS_ALLOC_FAILED:
      return "CUDNN_STATUS_ALLOC_FAILED";
    case CUDNN_STATUS_BAD_PARAM:
      return "CUDNN_STATUS_BAD_PARAM";
    case CUDNN_STATUS_INTERNAL_ERROR:
      return "CUDNN_STATUS_INTERNAL_ERROR";
    case CUDNN_STATUS_INVALID_VALUE:
      return "CUDNN_STATUS_INVALID_VALUE";
    case CUDNN_STATUS_ARCH_MISMATCH:
      return "CUDNN_STATUS_ARCH_MISMATCH";
    case CUDNN_STATUS_MAPPING_ERROR:
      return "CUDNN_STATUS_MAPPING_ERROR";
    case CUDNN_STATUS_EXECUTION_FAILED:
      return "CUDNN_STATUS_EXECUTION_FAILED";
    case CUDNN_STATUS_NOT_SUPPORTED:
      return "CUDNN_STATUS_NOT_SUPPORTED";
    case CUDNN_STATUS_LICENSE_ERROR:
      return "CUDNN_STATUS_LICENSE_ERROR";
#if CUDNN_VERSION_MIN(6, 0, 0)
    case CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING:
      return "CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING";
#endif
#if CUDNN_VERSION_MIN(7, 0, 0)
    case CUDNN_STATUS_RUNTIME_IN_PROGRESS:
      return "CUDNN_STATUS_RUNTIME_IN_PROGRESS";
    case CUDNN_STATUS_RUNTIME_FP_OVERFLOW:
      return "CUDNN_STATUS_RUNTIME_FP_OVERFLOW";
#endif
  }
  return "Unknown cudnn status";
}

namespace caffe {

namespace cudnn {

template <typename Dtype> class dataType;
template<> class dataType<float>  {
 public:
  static const cudnnDataType_t type = CUDNN_DATA_FLOAT;
  static float oneval, zeroval;
  static const void *one, *zero;
};
template<> class dataType<double> {
 public:
  static const cudnnDataType_t type = CUDNN_DATA_DOUBLE;
  static double oneval, zeroval;
  static const void *one, *zero;
};

template <typename Dtype>
inline void createTensor4dDesc(cudnnTensorDescriptor_t* desc) {
  CUDNN_CHECK(cudnnCreateTensorDescriptor(desc));
}

template <typename Dtype>
inline void setTensor4dDesc(cudnnTensorDescriptor_t* desc,
    int n, int c, int h, int w,
    int stride_n, int stride_c, int stride_h, int stride_w) {
  CUDNN_CHECK(cudnnSetTensor4dDescriptorEx(*desc, dataType<Dtype>::type,
        n, c, h, w, stride_n, stride_c, stride_h, stride_w));
}

template <typename Dtype>
inline void setTensor4dDesc(cudnnTensorDescriptor_t* desc,
    int n, int c, int h, int w) {
  const int stride_w = 1;
  const int stride_h = w * stride_w;
  const int stride_c = h * stride_h;
  const int stride_n = c * stride_c;
  setTensor4dDesc<Dtype>(desc, n, c, h, w,
                         stride_n, stride_c, stride_h, stride_w);
}

template <typename Dtype>
inline void createFilterDesc(cudnnFilterDescriptor_t* desc,
    int n, int c, int h, int w) {
  CUDNN_CHECK(cudnnCreateFilterDescriptor(desc));
#if CUDNN_VERSION_MIN(5, 0, 0)
  CUDNN_CHECK(cudnnSetFilter4dDescriptor(*desc, dataType<Dtype>::type,
      CUDNN_TENSOR_NCHW, n, c, h, w));
#else
  CUDNN_CHECK(cudnnSetFilter4dDescriptor_v4(*desc, dataType<Dtype>::type,
      CUDNN_TENSOR_NCHW, n, c, h, w));
#endif
}

template <typename Dtype>
inline void createConvolutionDesc(cudnnConvolutionDescriptor_t* conv) {
  CUDNN_CHECK(cudnnCreateConvolutionDescriptor(conv));
}

template <typename Dtype>
inline void setConvolutionDesc(cudnnConvolutionDescriptor_t* conv,
    cudnnTensorDescriptor_t bottom, cudnnFilterDescriptor_t filter,
    int pad_h, int pad_w, int stride_h, int stride_w) {
#if CUDNN_VERSION_MIN(6, 0, 0)
  CUDNN_CHECK(cudnnSetConvolution2dDescriptor(*conv,
      pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION,
      dataType<Dtype>::type));
#else
    CUDNN_CHECK(cudnnSetConvolution2dDescriptor(*conv,
      pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION));
#endif
}

template <typename Dtype>
inline void createPoolingDesc(cudnnPoolingDescriptor_t* pool_desc,
    PoolingParameter_PoolMethod poolmethod, cudnnPoolingMode_t* mode,
    int h, int w, int pad_h, int pad_w, int stride_h, int stride_w) {
  switch (poolmethod) {
  case PoolingParameter_PoolMethod_MAX:
    *mode = CUDNN_POOLING_MAX;
    break;
  case PoolingParameter_PoolMethod_AVE:
    *mode = CUDNN_POOLING_AVERAGE_COUNT_INCLUDE_PADDING;
    break;
  default:
    LOG(FATAL) << "Unknown pooling method.";
  }
  CUDNN_CHECK(cudnnCreatePoolingDescriptor(pool_desc));
#if CUDNN_VERSION_MIN(5, 0, 0)
  CUDNN_CHECK(cudnnSetPooling2dDescriptor(*pool_desc, *mode,
        CUDNN_PROPAGATE_NAN, h, w, pad_h, pad_w, stride_h, stride_w));
#else
  CUDNN_CHECK(cudnnSetPooling2dDescriptor_v4(*pool_desc, *mode,
        CUDNN_PROPAGATE_NAN, h, w, pad_h, pad_w, stride_h, stride_w));
#endif
}

template <typename Dtype>
inline void createActivationDescriptor(cudnnActivationDescriptor_t* activ_desc,
    cudnnActivationMode_t mode) {
  CUDNN_CHECK(cudnnCreateActivationDescriptor(activ_desc));
  CUDNN_CHECK(cudnnSetActivationDescriptor(*activ_desc, mode,
                                           CUDNN_PROPAGATE_NAN, Dtype(0)));
}

}  // namespace cudnn

}  // namespace caffe

#endif  // USE_CUDNN
#endif  // CAFFE_UTIL_CUDNN_H_

编译成功后，进入/home/xxx/ENet/enet_weights_zoo/，执行

$ sudo chmod a+x cityscapes_weights.sh
$ sh cityscapes_weights.sh

得到cityscapes_weights.caffemodel和cityscapes_weights_before_bn_merge.caffemodel。

（5）执行vision_segment_enet_detect节点

$ cd ~/autoware.ai
$ source install/setup.bash
$ roslaunch vision_segment_enet_detect vision_segment_enet_detect.launch

 如果没有image_segmenter_enet.launch文件，删除build和install里面的image_segmenter_enet,单独编译vision_segment_enet_detect。

$ cd ~/autoware.ai 
$ AUTOWARE_COMPILE_WITH_CUDA=1 colcon build --cmake-args -DCMAKE_BUILD_TYPE=Release --packages-select vision_segment_enet_detect

打开moriyama数据集测试。

如果出现：libcaffe.so.1.0.0-rc3: cannot open shared object file: No such file or directory错误

需要执行

$ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:~/ENet/caffe-enet/distribute/lib
$ roslaunch vision_segment_enet_detect vision_segment_enet_detect.launch

原创博文，转载请标明出处。