Signal Processing#

Sub-package containing basic signal and tensor manipulation routines.

Basic routines include tensor resizing (crop and pad), resampling (up- and downsampling), filtering, wavelet and low rank decompsition.

Resize and resampling#

deepmr.resize

Resize with zero-padding or cropping.

deepmr.resample

Resample a n-dimensional signal.

Filtering#

deepmr.fermi

Build a n-dimensional Fermi filter.

Tensor folding and unfolding#

deepmr.tensor2patches

View tensor as overlapping hyperectangular patches, with a given stride.

deepmr.patches2tensor

Accumulate patches into a tensor.

Wavelet decomposition#

deepmr.fwt

Forward wavelet transform.

deepmr.iwt

Inverse wavelet transform.

Signal compression#

deepmr.rss

Perform root sum-of-squares combination of a signal.

deepmr.svd

Perform SVD compression of a signal.