tiny_cnn.rar_CNN C_tiny-cnn_深度学习 图像

/*tiny-cnn训练*/ //#pragma warning(disable : 4996) #include <iostream> #include "tiny_cnn/tiny_cnn.h" #include <opencv2/opencv.hpp> #include<strstream> using namespace std; using namespace cv; using namespace tiny_cnn; using namespace tiny_cnn::activation; //**********************************************************************************// //定义全局变量 const int TRNUM = 200; //训练样本的个数 const int TENUM = 50; //测试样本的个数 const int C = 10; //分类的个数 //**********************************************************************************// //定义网络结构 void construct_net(network<mse, adagrad>& nn) { // connection table [Y.Lecun, 1998 Table.1] #define O true #define X false static const bool tbl[] = { O, X, X, X, O, O, O, X, X, O, O, O, O, X, O, O, O, O, X, X, X, O, O, O, X, X, O, O, O, O, X, O, O, O, O, X, X, X, O, O, O, X, X, O, X, O, O, O, X, O, O, O, X, X, O, O, O, O, X, X, O, X, O, O, X, X, O, O, O, X, X, O, O, O, O, X, O, O, X, O, X, X, X, O, O, O, X, X, O, O, O, O, X, O, O, O }; #undef O #undef X // construct nets nn << convolutional_layer<tan_h>(32, 32, 5, 1, 6) // C1, 1@32x32-in, 6@28x28-out << average_pooling_layer<tan_h>(28, 28, 6, 2) // S2, 6@28x28-in, 6@14x14-out << convolutional_layer<tan_h>(14, 14, 5, 6, 16, connection_table(tbl, 6, 16)) // C3, 6@14x14-in, 16@10x10-in << average_pooling_layer<tan_h>(10, 10, 16, 2) // S4, 16@10x10-in, 16@5x5-out << convolutional_layer<tan_h>(5, 5, 5, 16, 120) // C5, 16@5x5-in, 120@1x1-out << fully_connected_layer<tan_h>(120, C); // F6, 120-in, 10-out } //**********************************************************************************// // convert image to vec_t void convert_image(const string& imagefilename, double scale, int w, int h, std::vector<vec_t>& data) { auto img = imread(imagefilename, IMREAD_GRAYSCALE); if (img.data == nullptr) return; // cannot open, or it's not an image //imshow("img", img); //cvWaitKey(0); cv::Mat_<uint8_t> resized; cv::resize(img, resized, cv::Size(w, h)); vec_t d; std::transform(resized.begin(), resized.end(), std::back_inserter(d), [=](uint8_t c) { return c * scale; }); data.push_back(d); } //int 转换string string int2string(int&i){ strstream ss; string str; ss << i; ss >> str; return str; } int main() { //【第一步】定义网络结构 specify loss-function and learning strategy std::cout << "load models..." << std::endl; network<mse, adagrad> nn; construct_net(nn); //**********************************************************************************// //【第二步】加载数据 //加载训练数据 std::vector<label_t> train_labels, test_labels; std::vector<vec_t> train_images, test_images; for (int i = 1; i <= TRNUM; i++){ string str0 = int2string(i); string trainpath = "./tiny_cnn/train0/num (" + str0 + ").jpg"; //训练集路径 //转换图像格式 convert_image(trainpath, 1.0, 32, 32, train_images); } cout << "cout << train_images.size():" << train_images.size() << endl; //加载训练数据 for (int i = 1; i <= TENUM; i++){ string str1 = int2string(i); string testpath = "./tiny_cnn/val0/num (" + str1 + ").bmp"; //测试集路径 convert_image(testpath, 1.0, 32, 32, test_images); } cout << "cout << test_images.size():" << test_images.size() << endl; //**********************************************************************************// //【第三步】加载标签 手工添加 标签从0开始 printf("加载训练集标签\n"); int Ci; //每一类训练样本的个数 for (size_t k = 0; k < C; k++){ // C 分类的个数 printf("第i类样本的个数："); scanf("%d", &Ci); for (size_t j = 1; j <= Ci; j++){ train_labels.push_back((label_t)k); } } cout << "train_labels" << train_labels.size() << endl; printf("/加载测试集标签\n"); int Ti; //每一类测试样本的个数 for (size_t k = 0; k < C; k++){ // C 分类的个数 printf("第i类样本的个数："); scanf("%d", &Ti); for (size_t j = 1; j <= Ti; j++){ test_labels.push_back((label_t)k); } } cout << "test_labels" << test_labels.size() << endl; //**********************************************************************************// //【第四步】训练 std::cout << "start training" << std::endl; progress_display disp(train_images.size());//进度显示,初始化对象disp timer t; int minibatch_size = 10; int num_epochs = 30; //迭代次数 nn.optimizer().alpha *= std::sqrt(minibatch_size); // create callback auto on_enumerate_epoch = [&](){ std::cout << t.elapsed() << "s elapsed." << std::endl; tiny_cnn::result res = nn.test(test_images, test_labels); std::cout << res.num_success << "/" << res.num_total << std::endl; disp.restart(train_images.size()); t.restart(); }; auto on_enumerate_minibatch = [&](){ disp += minibatch_size; }; // training nn.train(train_images, //训练数据 train_labels, //标签 minibatch_size, num_epochs, on_enumerate_minibatch, on_enumerate_epoch); std::cout << "end training." << std::endl; // test and show results nn.test(test_images, test_labels).print_detail(std::cout); // 保存训练好的分类器 std::ofstream ofs("LeNet-weights"); ofs << nn; system("pause"); }