:jLLddlmZddlmZmZmZmZddlZddlmZddl m Z ddl m Z m Z mZddlmZddlmZdd lmZdd lmZmZesd d gZGd deZGddeZy))Sequence)AnyListOptionalUnionN)Tensor)Literal)_multiscale_ssim_update_ssim_check_inputs _ssim_update)Metric) dim_zero_cat)_MATPLOTLIB_AVAILABLE)_AX_TYPE_PLOT_OUT_TYPEz%StructuralSimilarityIndexMeasure.plotz/MultiScaleStructuralSimilarityIndexMeasure.plotcteZdZUdZdZeed<dZeed<dZeed<dZ e ed<d Z e ed <e e ed <e e ed < d dedee ee fdeeeefdeddeee ee e ffde de dedededd ffd Zd e d e dd fdZdee ee e fffdZ d!deee ee fdeedefdZxZS)" StructuralSimilarityIndexMeasurea4 Compute Structural Similarity Index Measure (SSIM_). As input to ``forward`` and ``update`` the metric accepts the following input - ``preds`` (:class:`~torch.Tensor`): Predictions from model - ``target`` (:class:`~torch.Tensor`): Ground truth values As output of `forward` and `compute` the metric returns the following output - ``ssim`` (:class:`~torch.Tensor`): if ``reduction!='none'`` returns float scalar tensor with average SSIM value over sample else returns tensor of shape ``(N,)`` with SSIM values per sample Args: preds: estimated image target: ground truth image gaussian_kernel: If ``True`` (default), a gaussian kernel is used, if ``False`` a uniform kernel is used sigma: Standard deviation of the gaussian kernel, anisotropic kernels are possible. Ignored if a uniform kernel is used kernel_size: the size of the uniform kernel, anisotropic kernels are possible. Ignored if a Gaussian kernel is used reduction: a method to reduce metric score over individual batch scores - ``'elementwise_mean'``: takes the mean - ``'sum'``: takes the sum - ``'none'`` or ``None``: no reduction will be applied data_range: the range of the data. If None, it is determined from the data (max - min). If a tuple is provided then the range is calculated as the difference and input is clamped between the values. k1: Parameter of SSIM. k2: Parameter of SSIM. return_full_image: If true, the full ``ssim`` image is returned as a second argument. Mutually exclusive with ``return_contrast_sensitivity`` return_contrast_sensitivity: If true, the constant term is returned as a second argument. The luminance term can be obtained with luminance=ssim/contrast Mutually exclusive with ``return_full_image`` kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Example: >>> import torch >>> from torchmetrics.image import StructuralSimilarityIndexMeasure >>> preds = torch.rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> ssim = StructuralSimilarityIndexMeasure(data_range=1.0) >>> ssim(preds, target) tensor(0.9219) Thigher_is_betteris_differentiableFfull_state_updateplot_lower_bound?plot_upper_boundpredstargetNgaussian_kernelsigma kernel_size reductionelementwise_meansumnoneN data_rangek1k2return_full_imagereturn_contrast_sensitivitykwargsreturnc  t |d i| d} || vrtd| d||dvr(|jdt j ddn|jdgd|jd t j dd| s|r|jd gd ||_||_||_||_ ||_ ||_ ||_ ||_ | |_y) Nr!$Argument `reduction` must be one of , but got r"r# similarityrr#defaultdist_reduce_fxtotal image_returncat)super__init__ ValueError add_statetorchtensorrrrr r%r&r'r(r)) selfrrrr r%r&r'r(r)r*valid_reduction __class__s l/home/mcse/projects/srt_converter/srt-converter-venv/lib/python3.12/site-packages/torchmetrics/image/ssim.pyr9z)StructuralSimilarityIndexMeasure.__init__Zs "6"C O +COCTT^_h^ijk k 3 3 NN<c1BSXN Y NN<DN I w S(9%P &*; NN>2eN L. &"$!2+F(c t||\}}t|||j|j|j|j |j |j|j|j }t|tr|\}}n|}|js |jr@t|jts td|jj|j dvrt|j"t$j&s td|xj"|j)z c_t|j*t$j&s td|xj*|j,dz c_yt|j"ts td|j"j|y)*Update state with predictions and targets.@Expected `self.image_return` to be a list when returning images.r/z9Expected `self.similarity` to be a Tensor for reductions.%Expected `self.total` to be a Tensor.rz>Expected `self.similarity` to be a list when reduction='none'.N)r r rrrr%r&r'r(r) isinstancetupler5list TypeErrorappendr r0r<rr#r4shape)r>rrsimilarity_packr0images rAupdatez'StructuralSimilarityIndexMeasure.updates^*5&9 v&    JJ    OO GG GG  " "  , ,   ou - / J(J  + +t/E/Ed//6 bcc    $ $U + >>8 8doou||< [\\ OOz~~/ /Odjj%,,7 GHH JJ%++a. (Jdoot4 `aa OO " ": .rBcz|jdk(rYt|jtr4t|jtr|j|jz }nt d|jdk(r2t|jts t d|j}n;t|jt rt|j}n t d|js |jr>t|jt rt|j}||fSt d|S)zCompute SSIM over state.r"z_Expected `self.similarity`and `self.total` to be of type Tensor for elementwise_mean reduction.r#r0r5s rAcomputez(StructuralSimilarityIndexMeasure.computes >>/ /$//62z$**f7U!__tzz9 u^^u $doov6 ^__J$//40)$//:  _``  + +t/E/E$++T2+D,=,=> |+ + bccrBvalaxc&|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 StructuralSimilarityIndexMeasure >>> preds = torch.rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> metric = StructuralSimilarityIndexMeasure(data_range=1.0) >>> metric.update(preds, target) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> import torch >>> from torchmetrics.image import StructuralSimilarityIndexMeasure >>> preds = torch.rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> metric = StructuralSimilarityIndexMeasure(data_range=1.0) >>> values = [ ] >>> for _ in range(10): ... values.append(metric(preds, target)) >>> fig_, ax_ = metric.plot(values) _plotr>rTrUs rAplotz%StructuralSimilarityIndexMeasure.plotXzz#r""rB) T? r"N{Gz?Q?FFNN)__name__ __module__ __qualname____doc__rbool__annotations__rrrfloatrrrrrintr rrHrr9rOrSrrrZ __classcell__r@s@rArrs/b"d!"t"#t#!e!!e! < L!%/213FXBF"',1%G%GUHUO+,%G3 -. %G BC %G U5%u *=#=>? %G %G %G %G&*%G%G %GN$/F$/F$/t$/LvuVV^'<<=:_c,#E&(6*:":;<,#IQRZI[,# ,#rBrcpeZdZUdZdZeed<dZeed<dZeed<dZ e ed<d Z e ed <e e ed <e e ed < d"dedeeeefdee ee fdeddeee ee e ffde de dee dfdeddedd ffd Zd e d e dd fdZde fdZ d#deee ee fd eedefd!ZxZS)$*MultiScaleStructuralSimilarityIndexMeasureaF Compute `MultiScaleSSIM`_, Multi-scale Structural Similarity Index Measure. This metric is is a generalization of Structural Similarity Index Measure by incorporating image details at different resolution scores. As input to ``forward`` and ``update`` the metric accepts the following input - ``preds`` (:class:`~torch.Tensor`): Predictions from model - ``target`` (:class:`~torch.Tensor`): Ground truth values As output of `forward` and `compute` the metric returns the following output - ``msssim`` (:class:`~torch.Tensor`): if ``reduction!='none'`` returns float scalar tensor with average MSSSIM value over sample else returns tensor of shape ``(N,)`` with SSIM values per sample Args: gaussian_kernel: If ``True`` (default), a gaussian kernel is used, if false a uniform kernel is used kernel_size: size of the gaussian kernel sigma: Standard deviation of the gaussian kernel reduction: a method to reduce metric score over labels. - ``'elementwise_mean'``: takes the mean - ``'sum'``: takes the sum - ``'none'`` or ``None``: no reduction will be applied data_range: the range of the data. If None, it is determined from the data (max - min). If a tuple is provided then the range is calculated as the difference and input is clamped between the values. The ``data_range`` must be given when ``dim`` is not None. k1: Parameter of structural similarity index measure. k2: Parameter of structural similarity index measure. betas: Exponent parameters for individual similarities and contrastive sensitivities returned by different image resolutions. normalize: When MultiScaleStructuralSimilarityIndexMeasure loss is used for training, it is desirable to use normalizes to improve the training stability. This `normalize` argument is out of scope of the original implementation [1], and it is adapted from https://github.com/jorge-pessoa/pytorch-msssim instead. kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Return: Tensor with Multi-Scale SSIM score Raises: ValueError: If ``kernel_size`` is not an int or a Sequence of ints with size 2 or 3. ValueError: If ``betas`` is not a tuple of floats with length 2. ValueError: If ``normalize`` is neither `None`, `ReLU` nor `simple`. Example: >>> from torch import rand >>> from torchmetrics.image import MultiScaleStructuralSimilarityIndexMeasure >>> preds = torch.rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> ms_ssim = MultiScaleStructuralSimilarityIndexMeasure(data_range=1.0) >>> ms_ssim(preds, target) tensor(0.9628) TrrFrrrrrrrNrrrr r!r%r&r'betas. normalize)relusimpleNr*r+c  t |di| d} || vrtd| d||dvr(|jdt j ddn|jdgd|jd t j ddt |ttfstd |t |tr-t|d vstd |Dstd |||_ ||_ ||_ ||_||_||_||_t |t$s tdt |t$rtd|Ds td||_| r| dvr td| |_y)Nr!r-r.r/r0rr#r1r4zRArgument `kernel_size` expected to be an sequence or an int, or a single int. Got )c3<K|]}t|tywN)rGrh).0kss rA zFMultiScaleStructuralSimilarityIndexMeasure.__init__..Ws5`bjS6I5`ztArgument `kernel_size` expected to be an sequence of size 2 or 3 where each element is an int, or a single int. Got z3Argument `betas` is expected to be of a type tuple.c3<K|]}t|tywru)rGrg)rvbetas rArxzFMultiScaleStructuralSimilarityIndexMeasure.__init__..gs/ZD 40G/Zryz5Argument `betas` is expected to be a tuple of floats.)rorpzNArgument `normalize` to be expected either `None` or one of 'relu' or 'simple'r7)r8r9r:r;r<r=rGrrhlenallrrrr r%r&r'rHrmrn) r>rrrr r%r&r'rmrnr*r?r@s rAr9z3MultiScaleStructuralSimilarityIndexMeasure.__init__8s "6"C O +COCTT^_h^ijk k 3 3 NN<c1BSXN Y NN<DN I w S(9%P;38depdqr  k8 ,  F *#5`T_5`2`((3}6  / &"$%'RS S eU #C/ZTY/Z,ZTU U *<<mn n"rBc "t||\}}t|||j|j|j|j |j |j|j|j }|jdvrAt|jts td|jj|nHt|jt s td|xj|j#z c_ t|j$t s td|xj$t'j(|j*d|j$j,|j$j.z c_y)rDr$NrRzPExpected `self.similarity` to be a Tensor for elementwise_mean or sum reduction.rFr)dtypedeviceN)r r rrrr%r&r'rmrnr rGr0rIrJrKrr#r4r<r=rLrr)r>rrr0s rArOz1MultiScaleStructuralSimilarityIndexMeasure.updatens*5&9 v,    JJ    OO GG GG JJ NN   >>^ +doot4 _`` OO " ": .doov6 rss OOz~~/ /O$**f-CD D ell5;;q>9I9IRVR\R\RcRcdd rBc|jdvr:t|jtrt |jSt d|jdk(r1t|jt r |jSt dt|jt r3t|jt r|j|jz St d)zCompute MS-SSIM over state.rrRr#rQzYExpected `self.similarity` and `self.total` to be Tensors for elementwise_mean reduction.)r rGr0rIrrJrr4)r>s rArSz2MultiScaleStructuralSimilarityIndexMeasure.computes >>^ +$//40#DOO44[\ \ >>U "$//62&Z[ [ doov .:djj&3Q??TZZ/ /sttrBrTrUc&|j||S)a"Plot 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 >>> from torch import rand >>> from torchmetrics.image import MultiScaleStructuralSimilarityIndexMeasure >>> preds = rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> metric = MultiScaleStructuralSimilarityIndexMeasure(data_range=1.0) >>> metric.update(preds, target) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> from torch import rand >>> from torchmetrics.image import MultiScaleStructuralSimilarityIndexMeasure >>> preds = rand([3, 3, 256, 256]) >>> target = preds * 0.75 >>> metric = MultiScaleStructuralSimilarityIndexMeasure(data_range=1.0) >>> values = [ ] >>> for _ in range(10): ... values.append(metric(preds, target)) >>> fig_, ax_ = metric.plot(values) rWrYs rArZz/MultiScaleStructuralSimilarityIndexMeasure.plotr[rB) Tr]r\r"Nr^r_)gǺ?g48EG?ga4?g??g9EGr?ror`)rarbrcrdrrerfrrrrgrrrrrhrr rrHrr9rOrSrrrZrirjs@rArlrls:x"d!"t"#t#!e!!e! < L!%13/2FXBF#K5;4#4#3 -.4#UHUO+, 4# BC 4# U5%u *=#=>? 4# 4# 4#UCZ 4#124#4# 4#leFeFete: u u_c,#E&(6*:":;<,#IQRZI[,# ,#rBrl)collections.abcrtypingrrrrr<rtyping_extensionsr "torchmetrics.functional.image.ssimr r r torchmetrics.metricr torchmetrics.utilities.datartorchmetrics.utilities.importsrtorchmetrics.utilities.plotrr__doctest_skip__rrlr7rBrArsW%-- %hh&4@@?ArsP#vP#fS#S#rB