In-Place Covariance Operators for Computer Vision