i%ddlmZddlmZmZmZddlZddlmZmZddl m Z m Z ddl m Z ddlmZddlmZmZesd gZGd d e Zy) )Sequence)AnyOptionalUnionN)Tensortensor)_psnrb_compute _psnrb_update)Metric)_MATPLOTLIB_AVAILABLE)_AX_TYPE_PLOT_OUT_TYPEz*PeakSignalNoiseRatioWithBlockedEffect.plotc eZdZUdZdZeed<dZeed<dZeed<e ed<e ed<e ed <e ed < dd e e e e e ffd e d ed dffd Zde de d dfdZd e fdZ ddee e ee fdeed efdZxZS)%PeakSignalNoiseRatioWithBlockedEffectaComputes `Peak Signal to Noise Ratio With Blocked Effect`_ (PSNRB). .. math:: \text{PSNRB}(I, J) = 10 * \log_{10} \left(\frac{\max(I)^2}{\text{MSE}(I, J)-\text{B}(I, J)}\right) Where :math:`\text{MSE}` denotes the `mean-squared-error`_ function. This metric is a modified version of PSNR that better supports evaluation of images with blocked artifacts, that oftens occur in compressed images. .. attention:: Metric only supports grayscale images. If you have RGB images, please convert them to grayscale first. As input to ``forward`` and ``update`` the metric accepts the following input - ``preds`` (:class:`~torch.Tensor`): Predictions from model of shape ``(N,1,H,W)`` - ``target`` (:class:`~torch.Tensor`): Ground truth values of shape ``(N,1,H,W)`` As output of `forward` and `compute` the metric returns the following output - ``psnrb`` (:class:`~torch.Tensor`): float scalar tensor with aggregated PSNRB value Args: data_range: the range of the data. If a tuple is provided then the range is calculated as the difference and input is clamped between the values. block_size: integer indication the block size kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Example: >>> from torch import rand >>> metric = PeakSignalNoiseRatioWithBlockedEffect(data_range=1.0) >>> preds = rand(2, 1, 10, 10) >>> target = rand(2, 1, 10, 10) >>> metric(preds, target) tensor(7.2893) Tis_differentiablehigher_is_betterFfull_state_updatesum_squared_errortotalbef data_range block_sizekwargsreturnNc t|d i|t|ts|dkr t d||_|j dtdd|j dtdd|j d tddttr1|j d tddz d fd |_ y|j d ttd d|_ y)Nz4Argument ``block_size`` should be a positive integerrgsum)defaultdist_reduce_fxrrrrmeanc@tj|ddS)Nrr)minmax)torchclamp)xrs m/Volumes/fast/ai/experiments/voice-extract-mac/.venv/lib/python3.12/site-packages/torchmetrics/image/psnrb.pyz@PeakSignalNoiseRatioWithBlockedEffect.__init__..\sQJqMzZ[})]) super__init__ isinstanceint ValueErrorr add_statertuple clamping_fnfloat)selfrrr __class__s ` r'r,z.PeakSignalNoiseRatioWithBlockedEffect.__init__Ks "6"*c*zA~ST T$ *F3KPUV wq %H ufSk%H j% ( NN< 1 ST 8U1VgmN n]D  NN<j8I1J[aN b#D r)predstargetc|j"|j|}|j|}t|||j\}}}|xj|z c_|xj|z c_|xj |z c_y)z*Update state with predictions and targets.N)r)r2r rrrr)r4r6r7rrnum_obss r'updatez,PeakSignalNoiseRatioWithBlockedEffect.updateasx    '$$U+E%%f-F*7vRVRaRa*b'3 "33 C g r)cnt|j|j|j|jS)z.Compute peak signal-to-noise ratio over state.)r rrrr)r4s r'computez-PeakSignalNoiseRatioWithBlockedEffect.computels%d44dhh DOO\\r)valaxc&|j||S)aPlot a single or multiple values from the metric. Args: val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results. If no value is provided, will automatically call `metric.compute` and plot that result. ax: An matplotlib axis object. If provided will add plot to that axis Returns: Figure and Axes object Raises: ModuleNotFoundError: If `matplotlib` is not installed .. plot:: :scale: 75 >>> # Example plotting a single value >>> import torch >>> from torchmetrics.image import PeakSignalNoiseRatioWithBlockedEffect >>> metric = PeakSignalNoiseRatioWithBlockedEffect(data_range=1.0) >>> metric.update(torch.rand(2, 1, 10, 10), torch.rand(2, 1, 10, 10)) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> import torch >>> from torchmetrics.image import PeakSignalNoiseRatioWithBlockedEffect >>> metric = PeakSignalNoiseRatioWithBlockedEffect(data_range=1.0) >>> values = [ ] >>> for _ in range(10): ... values.append(metric(torch.rand(2, 1, 10, 10), torch.rand(2, 1, 10, 10))) >>> fig_, ax_ = metric.plot(values) )_plot)r4r=r>s r'plotz*PeakSignalNoiseRatioWithBlockedEffect.plotpsPzz#r""r)))NN)__name__ __module__ __qualname____doc__rbool__annotations__rrrrr3r1r.rr,r:r<rrr rrA __classcell__)r5s@r'rrs"H#t"!d!#t# M K $%ue|!445$$ $  $, F F t ]] _c(#E&(6*:":;<(#IQRZI[(# (#r)r)collections.abcrtypingrrrr$rr#torchmetrics.functional.image.psnrbr r torchmetrics.metricr torchmetrics.utilities.importsr torchmetrics.utilities.plotr r__doctest_skip__rr*r)r'rQs<%'' M&@@DE{#F{#r)