riecovest.covariance_estimation.est_gevd

riecovest.covariance_estimation.est_gevd(Ry, Rn, rank, est_noise_cov=False)

Signal and noise covariance estimation using the GEVD method.

Parameters:

  • Ry (ndarray of shape (dim, dim) or (num_freqs, dim, dim)) – Sample covariance matrix of the noisy signal.

  • Rn (ndarray of shape (dim, dim) or (num_freqs, dim, dim)) – Sample covariance matrix of the noise.

  • rank (int) – Rank of the signal covariance matrix. Must be less than or equal to dim.

  • est_noise_cov (bool, optional) – If True, the noise covariance matrix is also estimated. Default is False.

Returns:

  • Rx_hat (ndarray of shape (dim, dim) or (num_freqs, dim, dim)) – Estimated signal covariance matrix.

  • Rv_hat (ndarray of shape (dim, dim) or (num_freqs, dim, dim)) – Estimated noise covariance matrix. Only returned if est_noise_cov is True.

References

[serizelLowrank2014] R. Serizel, M. Moonen, B. V. Dijk, and J. Wouters, “Low-rank approximation based multichannel Wiener filter algorithms for noise reduction with application in cochlear implants,” IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 22, no. 4, pp. 785–799, Apr. 2014, doi: 10.1109/TASLP.2014.2304240. [brunnstroemRobust2024] J. Brunnström, M. Moonen, and F. Elvander, “Robust signal and noise covariance matrix estimation using Riemannian optimization,” presented at the European Signal Processing Conference (EUSIPCO), Sep. 2024