将%caffe_root%\data\cifar10\get_cifar10.sh转换成linux格式,并用cygwin运行(事先装好wget)。自动下载并解压,得到batches.meta.txt,data_batch_1.bin,data_batch_2.bin,data_batch_3.bin,data_batch_4.bin,data_batch_5.bin,test_batch.bin。batches为标签对应的类别名;data_batch_x.bin为第x批10000张的训练图片,共5批;test_batch.bin为10000张测试图片。 创建%caffe_root%\data\cifar10\cifar10_images\test\data_batch_x(x为1-5共5个文件夹)及%caffe_root%\data\mnist\mnist_images\test_batch文件夹,以保存转换后的图片 编写如下python脚本,并运行。在对应文件夹下得到图片文件。 cifar10_data_root = r'd:/caffe/data/cifar10/' import struct import numpy as np from PIL import Image def load_CIFAR_Labels(filename): with open(filename, 'r') as f: lines = [] for line in f.readlines(): lines.append(line.strip('\n&