Calculating mean and std of images in a dataset in Fast AI

fastai
1

Hi All,

Is there a better way to compute the mean and standard deviation of all images in the dataset using some Fast AI convenient functions?

I am trying the following approach:

images = torch.stack([tensor(Image.open(o)).float()/255 for o in path.ls().sorted()])
mean = images.mean(axis=[0,1,2])

But it is incredibly slow.

I will appreciate your guidance.

Thanks and Kind Regards,
Bilal