TensorFlow MRI 0.19.0
Release 0.19.0
This release bumps the supported TensorFlow version to 2.9.
Major Features and Improvements
-
tfmri.image
:- Added new arguments
batch_dims
andimage_dims
to
image_gradients
,gradient_magnitude
,psnr
,ssim
and
ssim_multiscale
. - Argument
rank
ofpsnr
,ssim
andssim_multiscale
is now
deprecated. To update, useimage_dims
instead. image_gradients
andgradient_magnitude
now support complex inputs.
- Added new arguments
-
tfmri.losses
:- Added new arguments
batch_dims
andimage_dims
to all image quality
losses. Argumentrank
is now deprecated (useimage_dims
instead). - Added new arguments
multichannel
andcomplex_part
to all image
quality losses. These mirror the behaviour of the corresponding metrics
and allows for single-channel inputs and extraction of parts from complex
numbers, respectively. - The canonical API names
StructuralSimilarityLoss
and
MultiscaleStructuralSimilarityLoss
have been
changed toSSIMLoss
andSSIMMultiscaleLoss
, respectively. The
old names are still available, but may be removed in the future.
- Added new arguments
-
tfmri.metrics
:- Image quality metrics can now accept complex inputs without also specifying
complex_part
, in which case the unmodified complex values will be passed
to the downstream function. This may not be supported for all metrics. - Added new arguments
batch_dims
andimage_dims
to all image quality
metrics. Argumentrank
is now deprecated (useimage_dims
instead). - The canonical API names
PeakSignalToNoiseRatio
,
StructuralSimilarity
andMultiscaleStructuralSimilarity
have been
changed toPSNR
,SSIM
andSSIMMultiscale
, respectively. The
old names are still available, but may be removed in the future.
- Image quality metrics can now accept complex inputs without also specifying
-
tfmri.recon
:- Added new argument
preserve_phase
totfmri.recon.pf
. This allows
the user to recover the phase as well as the magnitude during partial
Fourier reconstruction. Argumentreturn_complex
has the same behaviour
and is now deprecated. - Added new aliases
adjoint
(foradj
),least_squares
(forlstsq
) andpartial_fourier
(forpf
). These are now the
canonical aliases, but the old ones will still be supported.
- Added new argument
-
tfmri.signal
:- Added new ops
dwt
andidwt
to compute the N-dimensional discrete
wavelet transform and its inverse, respectively.
- Added new ops
-
tfmri.plot
:- Added new argument
norm
toimage_sequence
,tiled_image
and
tiled_image_sequence
. This allows the user to specify the scaling
to be applied before the colormap.
- Added new argument
Bug Fixes and Other Changes
- Fixed a bug with k-space weighting in homodyne detection method of
tfmri.recon.partial_fourier
. - Like core TensorFlow, we now compile with
_GLIBCXX_USE_CXX11_ABI=1
. - Like core TensorFlow, Python wheels now conform to
manylinux2014
, an
upgrade frommanylinux2010
.