pytransform3d.uncertainty.estimate_gaussian_rotation_matrix_from_samples

pytransform3d.uncertainty.estimate_gaussian_rotation_matrix_from_samples(samples)

Estimate Gaussian distribution over rotations from samples.

Computes the Fréchet mean of the samples and the covariance in tangent space (exponential coordinates of rotation / rotation vectors) using an unbiased estimator as outlines by Eade [1].

Parameters:

samplesarray-like, shape (n_samples, 3, 3)

Sampled rotations represented by rotation matrices.

Returns:

meanarray, shape (3, 3)

Mean of the Gaussian distribution as rotation matrix.

covarray, shape (3, 3)

Covariance of the Gaussian distribution in exponential coordinates.

See also

frechet_mean

Algorithm used to compute the mean of the Gaussian.

References

[1]

Eade, E. (2017). Lie Groups for 2D and 3D Transformations. https://ethaneade.com/lie.pdf