Caffe图片训练分类研究、深度学习图片分类
转载请注明：http://blog.csdn.net/forest_world
一、NSFW研究

1、安装Docker

http://www.linuxidc.com/Linux/2014-08/105656.htm

安装Docker使用apt-get命令:
$ apt-get install docker.io
创建软连接
ln -sf /usr/bin/docker.io /usr/local/bin/docker

sudo service docker stop
sudo service docker start 

2、

sudo docker build -t caffe:cpu https://raw.githubusercontent.com/BVLC/caffe/master/docker/standalone/cpu/Dockerfile
Step 0 : FROM ubuntu:14.04
 ---> 35b394a6f7a2
Step 1 : MAINTAINER caffe-maint@googlegroups.com
 ---> Using cache
 ---> ca50125d0951
Step 2 : RUN apt-get update && apt-get install -y --no-install-recommends         build-essential         cmake         git         wget         libatlas-base-dev         libboost-all-dev         libgflags-dev         libgoogle-glog-dev         libhdf5-serial-dev         libleveldb-dev         liblmdb-dev         libopencv-dev         libprotobuf-dev         libsnappy-dev         protobuf-compiler         python-dev         python-numpy         python-pip         python-scipy &&     rm -rf /var/lib/apt/lists/*
 ---> Running in d6856e1b4740
Ign http://archive.ubuntu.com trusty InRelease
Get:1 http://archive.ubuntu.com trusty-updates InRelease [65.9 kB]
Get:2 http://archive.ubuntu.com trusty-security InRelease [65.9 kB]
Get:3 http://archive.ubuntu.com trusty Release.gpg [933 B]
Get:4 http://archive.ubuntu.com trusty-updates/main Sources [474 kB]
Get:5 http://archive.ubuntu.com trusty-updates/main Sources [474 kB]
Get:6 http://archive.ubuntu.com trusty-updates/restricted Sources [5247 B]
Get:7 http://archive.ubuntu.com trusty-updates/universe Sources [209 kB]
Get:8 http://archive.ubuntu.com trusty-updates/main amd64 Packages [1131 kB]

......

Removing intermediate container d3643cce1d7e
Step 7 : ENV PYCAFFE_ROOT $CAFFE_ROOT/python
 ---> Running in e4e4019889f8
 ---> e982c669b99b
Removing intermediate container e4e4019889f8
Step 8 : ENV PYTHONPATH $PYCAFFE_ROOT:$PYTHONPATH
 ---> Running in a9ee4331bbe8
 ---> 8a1867b64b5c
Removing intermediate container a9ee4331bbe8
Step 9 : ENV PATH $CAFFE_ROOT/build/tools:$PYCAFFE_ROOT:$PATH
 ---> Running in bc2a271a95bd
 ---> 864daab5c633
Removing intermediate container bc2a271a95bd
Step 10 : RUN echo "$CAFFE_ROOT/build/lib" >> /etc/ld.so.conf.d/caffe.conf && ldconfig
 ---> Running in d0af6f3e69ea
 ---> fa8b1e810492
Removing intermediate container d0af6f3e69ea
Step 11 : WORKDIR /workspace
 ---> Running in ab94152a0a18
 ---> 49ffbf2d8fef
Removing intermediate container ab94152a0a18
Successfully built 49ffbf2d8fef

3、
git clone https://github.com/yahoo/open_nsfw
$ cd open_nsfw

@ubuntu:~$ git clone https://github.com/yahoo/open_nsfw
Cloning into 'open_nsfw'...
remote: Counting objects: 31, done.
remote: Compressing objects: 100% (20/20), done.
Unpacking objects:  32% (10/31) 

4、

I1012 05:17:23.226325     1 net.cpp:228] relu_stage0_block0 does not need backward computation.
I1012 05:17:23.226327     1 net.cpp:228] eltwise_stage0_block0 does not need backward computation.
I1012 05:17:23.226331     1 net.cpp:228] scale_stage0_block0_branch2c does not need backward computation.
I1012 05:17:23.226333     1 net.cpp:228] bn_stage0_block0_branch2c does not need backward computation.
I1012 05:17:23.226336     1 net.cpp:228] conv_stage0_block0_branch2c does not need backward computation.
I1012 05:17:23.226339     1 net.cpp:228] relu_stage0_block0_branch2b does not need backward computation.
I1012 05:17:23.226342     1 net.cpp:228] scale_stage0_block0_branch2b does not need backward computation.
I1012 05:17:23.226346     1 net.cpp:228] bn_stage0_block0_branch2b does not need backward computation.
I1012 05:17:23.226348     1 net.cpp:228] conv_stage0_block0_branch2b does not need backward computation.
I1012 05:17:23.226351     1 net.cpp:228] relu_stage0_block0_branch2a does not need backward computation.
I1012 05:17:23.226354     1 net.cpp:228] scale_stage0_block0_branch2a does not need backward computation.
I1012 05:17:23.226356     1 net.cpp:228] bn_stage0_block0_branch2a does not need backward computation.
I1012 05:17:23.226359     1 net.cpp:228] conv_stage0_block0_branch2a does not need backward computation.
I1012 05:17:23.226362     1 net.cpp:228] scale_stage0_block0_proj_shortcut does not need backward computation.
I1012 05:17:23.226366     1 net.cpp:228] bn_stage0_block0_proj_shortcut does not need backward computation.
I1012 05:17:23.226368     1 net.cpp:228] conv_stage0_block0_proj_shortcut does not need backward computation.
I1012 05:17:23.226372     1 net.cpp:228] pool1_pool1_0_split does not need backward computation.
I1012 05:17:23.226374     1 net.cpp:228] pool1 does not need backward computation.
I1012 05:17:23.226378     1 net.cpp:228] relu_1 does not need backward computation.
I1012 05:17:23.226380     1 net.cpp:228] scale_1 does not need backward computation.
I1012 05:17:23.226383     1 net.cpp:228] bn_1 does not need backward computation.
I1012 05:17:23.226387     1 net.cpp:228] conv_1 does not need backward computation.
I1012 05:17:23.226389     1 net.cpp:228] data does not need backward computation.
I1012 05:17:23.226392     1 net.cpp:270] This network produces output prob
I1012 05:17:23.226526     1 net.cpp:283] Network initialization done.
I1012 05:17:23.277700     1 upgrade_proto.cpp:77] Attempting to upgrade batch norm layers using deprecated params: nsfw_model/resnet_50_1by2_nsfw.caffemodel
I1012 05:17:23.277819     1 upgrade_proto.cpp:80] Successfully upgraded batch norm layers using deprecated params.
I1012 05:17:23.283418     1 net.cpp:761] Ignoring source layer loss
NSFW score:   0.000410715758335

二、

#include <caffe/caffe.hpp>
#ifdef USE_OPENCV
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#endif  // USE_OPENCV
#include <algorithm>
#include <iosfwd>
#include <memory>
#include <string>
#include <utility>
#include <vector>

#ifdef USE_OPENCV
using namespace caffe;  // NOLINT(build/namespaces)
using std::string;

/* Pair (label, confidence) representing a prediction. */
typedef std::pair<string, float> Prediction;

class Classifier {
 public:
  Classifier(const string& model_file,
             const string& trained_file,
             const string& mean_file,
             const string& label_file);

  std::vector<Prediction> Classify(const cv::Mat& img, int N = 5);

 private:
  void SetMean(const string& mean_file);

  std::vector<float> Predict(const cv::Mat& img);

  void WrapInputLayer(std::vector<cv::Mat>* input_channels);

  void Preprocess(const cv::Mat& img,
                  std::vector<cv::Mat>* input_channels);

 private:
  shared_ptr<Net<float> > net_;
  cv::Size input_geometry_;
  int num_channels_;
  cv::Mat mean_;
  std::vector<string> labels_;
};

Classifier::Classifier(const string& model_file,
                       const string& trained_file,
                       const string& mean_file,
                       const string& label_file) {
#ifdef CPU_ONLY
  Caffe::set_mode(Caffe::CPU);
#else
  Caffe::set_mode(Caffe::GPU);
#endif

  /* Load the network. */
  net_.reset(new Net<float>(model_file, TEST));
  net_->CopyTrainedLayersFrom(trained_file);

  CHECK_EQ(net_->num_inputs(), 1) << "Network should have exactly one input.";
  CHECK_EQ(net_->num_outputs(), 1) << "Network should have exactly one output.";

  Blob<float>* input_layer = net_->input_blobs()[0];
  num_channels_ = input_layer->channels();
  CHECK(num_channels_ == 3 || num_channels_ == 1)
    << "Input layer should have 1 or 3 channels.";
  input_geometry_ = cv::Size(input_layer->width(), input_layer->height());

  /* Load the binaryproto mean file. */
  SetMean(mean_file);

  /* Load labels. */
  std::ifstream labels(label_file.c_str());
  CHECK(labels) << "Unable to open labels file " << label_file;
  string line;
  while (std::getline(labels, line))
    labels_.push_back(string(line));

  Blob<float>* output_layer = net_->output_blobs()[0];
  CHECK_EQ(labels_.size(), output_layer->channels())
    << "Number of labels is different from the output layer dimension.";
}

static bool PairCompare(const std::pair<float, int>& lhs,
                        const std::pair<float, int>& rhs) {
  return lhs.first > rhs.first;
}

/* Return the indices of the top N values of vector v. */
static std::vector<int> Argmax(const std::vector<float>& v, int N) {
  std::vector<std::pair<float, int> > pairs;
  for (size_t i = 0; i < v.size(); ++i)
    pairs.push_back(std::make_pair(v[i], static_cast<int>(i)));
  std::partial_sort(pairs.begin(), pairs.begin() + N, pairs.end(), PairCompare);

  std::vector<int> result;
  for (int i = 0; i < N; ++i)
    result.push_back(pairs[i].second);
  return result;
}

/* Return the top N predictions. */
std::vector<Prediction> Classifier::Classify(const cv::Mat& img, int N) {
  std::vector<float> output = Predict(img);

  N = std::min<int>(labels_.size(), N);
  std::vector<int> maxN = Argmax(output, N);
  std::vector<Prediction> predictions;
  for (int i = 0; i < N; ++i) {
    int idx = maxN[i];
    predictions.push_back(std::make_pair(labels_[idx], output[idx]));
  }

  return predictions;
}

/* Load the mean file in binaryproto format. */
void Classifier::SetMean(const string& mean_file) {
  BlobProto blob_proto;
  ReadProtoFromBinaryFileOrDie(mean_file.c_str(), &blob_proto);

  /* Convert from BlobProto to Blob<float> */
  Blob<float> mean_blob;
  mean_blob.FromProto(blob_proto);
  CHECK_EQ(mean_blob.channels(), num_channels_)
    << "Number of channels of mean file doesn't match input layer.";

  /* The format of the mean file is planar 32-bit float BGR or grayscale. */
  std::vector<cv::Mat> channels;
  float* data = mean_blob.mutable_cpu_data();
  for (int i = 0; i < num_channels_; ++i) {
    /* Extract an individual channel. */
    cv::Mat channel(mean_blob.height(), mean_blob.width(), CV_32FC1, data);
    channels.push_back(channel);
    data += mean_blob.height() * mean_blob.width();
  }

  /* Merge the separate channels into a single image. */
  cv::Mat mean;
  cv::merge(channels, mean);

  /* Compute the global mean pixel value and create a mean image
   * filled with this value. */
  cv::Scalar channel_mean = cv::mean(mean);
  mean_ = cv::Mat(input_geometry_, mean.type(), channel_mean);
}

std::vector<float> Classifier::Predict(const cv::Mat& img) {
  Blob<float>* input_layer = net_->input_blobs()[0];
  input_layer->Reshape(1, num_channels_,
                       input_geometry_.height, input_geometry_.width);
  /* Forward dimension change to all layers. */
  net_->Reshape();

  std::vector<cv::Mat> input_channels;
  WrapInputLayer(&input_channels);

  Preprocess(img, &input_channels);

  net_->Forward();

  /* Copy the output layer to a std::vector */
  Blob<float>* output_layer = net_->output_blobs()[0];
  const float* begin = output_layer->cpu_data();
  const float* end = begin + output_layer->channels();
  return std::vector<float>(begin, end);
}

/* Wrap the input layer of the network in separate cv::Mat objects
 * (one per channel). This way we save one memcpy operation and we
 * don't need to rely on cudaMemcpy2D. The last preprocessing
 * operation will write the separate channels directly to the input
 * layer. */
void Classifier::WrapInputLayer(std::vector<cv::Mat>* input_channels) {
  Blob<float>* input_layer = net_->input_blobs()[0];

  int width = input_layer->width();
  int height = input_layer->height();
  float* input_data = input_layer->mutable_cpu_data();
  for (int i = 0; i < input_layer->channels(); ++i) {
    cv::Mat channel(height, width, CV_32FC1, input_data);
    input_channels->push_back(channel);
    input_data += width * height;
  }
}

void Classifier::Preprocess(const cv::Mat& img,
                            std::vector<cv::Mat>* input_channels) {
  /* Convert the input image to the input image format of the network. */
  cv::Mat sample;
  if (img.channels() == 3 && num_channels_ == 1)
    cv::cvtColor(img, sample, cv::COLOR_BGR2GRAY);
  else if (img.channels() == 4 && num_channels_ == 1)
    cv::cvtColor(img, sample, cv::COLOR_BGRA2GRAY);
  else if (img.channels() == 4 && num_channels_ == 3)
    cv::cvtColor(img, sample, cv::COLOR_BGRA2BGR);
  else if (img.channels() == 1 && num_channels_ == 3)
    cv::cvtColor(img, sample, cv::COLOR_GRAY2BGR);
  else
    sample = img;

  cv::Mat sample_resized;
  if (sample.size() != input_geometry_)
    cv::resize(sample, sample_resized, input_geometry_);
  else
    sample_resized = sample;

  cv::Mat sample_float;
  if (num_channels_ == 3)
    sample_resized.convertTo(sample_float, CV_32FC3);
  else
    sample_resized.convertTo(sample_float, CV_32FC1);

  cv::Mat sample_normalized;
  cv::subtract(sample_float, mean_, sample_normalized);

  /* This operation will write the separate BGR planes directly to the
   * input layer of the network because it is wrapped by the cv::Mat
   * objects in input_channels. */
  cv::split(sample_normalized, *input_channels);

  CHECK(reinterpret_cast<float*>(input_channels->at(0).data)
        == net_->input_blobs()[0]->cpu_data())
    << "Input channels are not wrapping the input layer of the network.";
}

int main(int argc, char** argv) {
  if (argc != 6) {
    std::cerr << "Usage: " << argv[0]
              << " deploy.prototxt network.caffemodel"
              << " mean.binaryproto labels.txt img.jpg" << std::endl;
    return 1;
  }

  ::google::InitGoogleLogging(argv[0]);

  string model_file   = argv[1];
  string trained_file = argv[2];
  string mean_file    = argv[3];
  string label_file   = argv[4];
  Classifier classifier(model_file, trained_file, mean_file, label_file);

  string file = argv[5];

  std::cout << "---------- Prediction for "
            << file << " ----------" << std::endl;

  cv::Mat img = cv::imread(file, -1);
  CHECK(!img.empty()) << "Unable to decode image " << file;
  std::vector<Prediction> predictions = classifier.Classify(img);

  /* Print the top N predictions. */
  for (size_t i = 0; i < predictions.size(); ++i) {
    Prediction p = predictions[i];
    std::cout << std::fixed << std::setprecision(4) << p.second << " - \""
              << p.first << "\"" << std::endl;
  }
}
#else
int main(int argc, char** argv) {
  LOG(FATAL) << "This example requires OpenCV; compile with USE_OPENCV.";
}
#endif  // USE_OPENCV

  string model_file   ("E:\\ cpp_classification\\caffe.prototxt");
  string trained_file ("E:\\ cpp_classification\\caffe.caffemodel");
  string mean_file    ("E:\\cpp_classification\\mean.binaryproto");
  string label_file   ("E:\\ cpp_classification\\labels.txt");
  Classifier classifier(model_file, trained_file, mean_file, label_file);

  string file ("E:\\ cpp_classification\\test.jpg");

参考学习资料：
http://m.blog.csdn.net/article/details?id=52443126 基于深度学习的人脸识别系统系列（Caffe+OpenCV+Dlib）——【一】如何在Visual Studio中像使用OpenCV一样使用Caffe
http://mp.weixin.qq.com/s?__biz=MzI1NTE4NTUwOQ==&mid=2650325557&idx=1&sn=362d476d3b3820ea56e4672369565e4f&chksm=f235a53fc5422c2939f76b7e8f5265333f3159b0ec4275fe733d27e7a03f17395b0460a318d2&mpshare=1&scene=1&srcid=1017Le0xZeDhioc9DxPIGNN9#wechat_redirect IJCAI16论文速读：Deep Learning论文选读（上）

http://www.cnblogs.com/carle-09/p/5779304.html 4 .caffe：train_val.prototxt、 solver.prototxt 、 deploy.prototxt（ 创建模型与编写配置文件）
http://blog.csdn.net/deeplearninglc007/article/details/40086503 使用Caffe对图片进行训练并分类的简单流程
http://blog.csdn.net/wang4959520/article/details/51841110 将train_val.prototxt 转换成deploy.prototxt
http://blog.csdn.net/hyman_yx/article/details/51732656 Caffe均值文件mean.binaryproto转mean.npy
http://blog.csdn.net/shakevincent/article/details/51694686微软Caffe编译
http://www.cnblogs.com/alexcai/p/5469436.html caffe简易上手指南（二）—— 训练我们自己的数据
http://www.aiuxian.com/article/p-1659539.html 深度学习–如何利用Caffe进行训练ImageNet网络
http://www.th7.cn/system/win/201602/153606.shtml caffe for windows 下使用caffemodel 实现cifar10的图像分类
http://blog.csdn.net/dcxhun3/article/details/52021296 用训练好的caffemodel来进行分类
http://neuralnetworksanddeeplearning.com/chap1.html CHAPTER 1 Using neural nets to recognize handwritten digits

http://www.cnblogs.com/shishupeng/p/5694775.html 深度卷积网络CNN与图像语义分割

本文地址：http://blog.csdn.net/forest_world