关于卷积神经网络，读文件的代码，卷积的代码（并行优化过的）、池化的代码。(About convolutional neural network, read the file's code, convolution code (parallel optimization), pool code.)