训练神经网络时出错

Question

我正在努力训练神经网络

            Ptr< ANN_MLP >  nn = ANN_MLP::create();
                nn->setActivationFunction(cv::ml::ANN_MLP::SIGMOID_SYM);
                nn->setTrainMethod(cv::ml::ANN_MLP::BACKPROP);
                nn->setBackpropMomentumScale(0.1);
                nn->setBackpropWeightScale(0.1);
                nn->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, (int)100000, 1e-6));

                //setting the NN layer size
                cv::Mat layers = cv::Mat(3, 1, CV_32SC1);
                layers.row(0) = cv::Scalar(4);
                layers.row(1) = cv::Scalar(10);
                layers.row(2) = cv::Scalar(16);
                //layers.row(3) = cv::Scalar(16);
                nn->setLayerSizes(layers);
                nn->train(train_data, ROW_SAMPLE, response);
                cout << "Hari";
                while (1){}
                waitKey(0);
                return 0;

trainData是105735 * 4矩阵，而响应是105735 * 16矩阵。

以下错误即将来临：                                             BTP.exe中0x00007FFA81ECA1C8处的未处理异常：Microsoft C ++异常：cv ::内存位置0x000000322AA5D820的异常。

错误是：

  

OpenCV错误：在cv :: ml中断言失败（（layout == ROW_SAMPLE＆amp;＆amp; responses.rows == nsamples）||（layout == COL_SAMPLE＆amp;＆amp; respon.cols == nsamples））： ：TrainDataImpl :: setData，文件C：\ builds \ master_PackSlave-win64-vc12-shared \ opencv \ modules \ ml \ src \ data.cpp，第290行