计算pytorch标准化（Normalize）所需要数据集的均值和
⽅差实例
pytorch做标准化利⽤transforms.Normalize(mean_vals, std_vals)，其中常⽤数据集的均值⽅差有：
if 'coco' in args.dataset:
mean_vals = [0.471, 0.448, 0.408]
std_vals = [0.234, 0.239, 0.242]
elif 'imagenet' in args.dataset:
mean_vals = [0.485, 0.456, 0.406]
std_vals = [0.229, 0.224, 0.225]
计算⾃⼰数据集图像像素的均值⽅差：
import numpy as np
import cv2
import random
# calculate means and std
train_txt_path = './train_'
CNum = 10000  # 挑选多少图⽚进⾏计算
img_h, img_w = 32, 32
imgs = np.zeros([img_w, img_h, 3, 1])
means, stdevs = [], []
with open(train_txt_path, 'r') as f:
lines = f.readlines()
random.shuffle(lines)  # shuffle , 随机挑选图⽚
for i in tqdm_notebook(range(CNum)):
img_path = os.path.join('./train', lines[i].rstrip().split()[0])
img = cv2.imread(img_path)
img = size(img, (img_h, img_w))
img = img[:, :, :, np.newaxis]
imgs = np.concatenate((imgs, img), axis=3)
#    print(i)
imgs = imgs.astype(np.float32)/255.
for i in tqdm_notebook(range(3)):
pixels = imgs[:,:,i,:].ravel() # 拉成⼀⾏
means.an(pixels))
numpy库是标准库吗stdevs.append(np.std(pixels))
# cv2 读取的图像格式为BGR，PIL/Skimage读取到的都是RGB不⽤转
print("normMean = {}".format(means))
print("normStd = {}".format(stdevs))
print('transforms.Normalize(normMean = {}, normStd = {})'.format(means, stdevs))
以上这篇计算pytorch标准化(Normalize)所需要数据集的均值和⽅差实例就是⼩编分享给⼤家的全部内容了，希望能给⼤家⼀个参考，也希望⼤家多多⽀持。