FID Feature Statistics

Problem

Compute the feature statistics used by FID: the mean vector and the covariance matrix (divided by N - 1) of a (num_samples, dim) feature matrix.

The signature is fid_feature_stats(features). Return [mean, covariance].

Hint

FID compares the mean and covariance of real and generated features.

case 1

in: [[[1,0],[0,1]]]

expected: [[0.5,0.5],[[0.5,-0.5],[-0.5,0.5]]]

case 2

in: [[[1,2],[3,4],[5,6]]]

expected: [[3,4],[[4,4],[4,4]]]

case 3

in: [[[0,0],[0,0]]]

expected: [[0,0],[[0,0],[0,0]]]

case 4

in: [[[2,1,0],[1,2,3]]]

expected: [[1.5,1.5,1.5],[[0.5,-0.5,-1.5],[-0.5,0.5,1.5],[-1.5,1.5,4.5]]]

Starter codePython

def fid_feature_stats(features):
    # Your code here
    pass

4 test cases · run your code against them in the editor