o .wi@sddlmZddlmZmZmZmZddlmZm Z ddl m Z ddl m Z mZddlmZddlmZddlmZdd lmZdd lmZmZesMd gZGd d d eZdS))Sequence)AnyListOptionalUnion)Tensortensor)Literal) _uqi_compute _uqi_update)Metric)rank_zero_warn) dim_zero_cat)_MATPLOTLIB_AVAILABLE)_AX_TYPE_PLOT_OUT_TYPEUniversalImageQualityIndex.plotc seZdZUdZdZeed<dZeed<dZeed<dZ e ed<d Z e ed <e e ed <e e ed <e ed <e ed<   d#deedee deddeddf fdd Zd e d e ddfddZde fddZ d$deee ee fd eedefd!d"ZZS)%UniversalImageQualityIndexaCompute Universal Image Quality Index (UniversalImageQualityIndex_). As input to ``forward`` and ``update`` the metric accepts the following input - ``preds`` (:class:`~torch.Tensor`): Predictions from model of shape ``(N,C,H,W)`` - ``target`` (:class:`~torch.Tensor`): Ground truth values of shape ``(N,C,H,W)`` As output of `forward` and `compute` the metric returns the following output - ``uiqi`` (:class:`~torch.Tensor`): if ``reduction!='none'`` returns float scalar tensor with average UIQI value over sample else returns tensor of shape ``(N,)`` with UIQI values per sample Args: kernel_size: size of the gaussian kernel sigma: Standard deviation of the gaussian kernel reduction: a method to reduce metric score over labels. - ``'elementwise_mean'``: takes the mean (default) - ``'sum'``: takes the sum - ``'none'`` or ``None``: no reduction will be applied kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Return: Tensor with UniversalImageQualityIndex score Example: >>> import torch >>> from torchmetrics.image import UniversalImageQualityIndex >>> preds = torch.rand([16, 1, 16, 16]) >>> target = preds * 0.75 >>> uqi = UniversalImageQualityIndex() >>> uqi(preds, target) tensor(0.9216) Tis_differentiablehigher_is_betterFfull_state_updateplot_lower_boundg?plot_upper_boundpredstargetsum_uqinumel r?r!elementwise_mean kernel_sizesigma reductionr"sumnoneNkwargsreturnNc stjdi||dvrtd|d|dus|dkr2td|jdgdd|jd gddn|jd td d d |jdtdd d ||_||_||_dS)Nr&zThe `reduction` zI is not valid. Valid options are `elementwise_mean`, `sum`, `none`, None.r(zMetric `UniversalImageQualityIndex` will save all targets and predictions in the buffer when using`reduction=None` or `reduction='none'. For large datasets, this may lead to a large memory footprint.rcat)defaultdist_reduce_fxrrrr')r-rr) super__init__ ValueErrorr add_staterr#r$r%)selfr#r$r%r) __class__r.S/home/ubuntu/sommelier/.venv/lib/python3.10/site-packages/torchmetrics/image/uqi.pyr0Ps   z#UniversalImageQualityIndex.__init__cCst||\}}|jdus|jdkr|j||j|dSt|||j|jdd}|j|7_|j }|j |d|d|d|jdd|d|jdd7_ dS) z*Update state with predictions and targets.Nr(r')r%r) r r%rappendrr r#r$rshaper)r3rr uqi_scorepsr.r.r6updatejs Jz!UniversalImageQualityIndex.updatecCsZ|jdks |jdurt|j}t|j}t|||j|j|jS|jdkr*|j|jS|jS)z&Compute explained variance over state.r(Nr") r%rrrr r#r$rr)r3rrr.r.r6computevs   z"UniversalImageQualityIndex.computevalaxcCs |||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 UniversalImageQualityIndex >>> preds = torch.rand([16, 1, 16, 16]) >>> target = preds * 0.75 >>> metric = UniversalImageQualityIndex() >>> metric.update(preds, target) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> import torch >>> from torchmetrics.image import UniversalImageQualityIndex >>> preds = torch.rand([16, 1, 16, 16]) >>> target = preds * 0.75 >>> metric = UniversalImageQualityIndex() >>> values = [ ] >>> for _ in range(10): ... values.append(metric(preds, target)) >>> fig_, ax_ = metric.plot(values) )_plot)r3r@rAr.r.r6plot~s ,r)rr r")NN)__name__ __module__ __qualname____doc__rbool__annotations__rrrfloatrrrrintr rr0r>r?rrrrrC __classcell__r.r.r4r6rsF  %        rN)collections.abcrtypingrrrrtorchrrtyping_extensionsr !torchmetrics.functional.image.uqir r torchmetrics.metricr torchmetrics.utilitiesr torchmetrics.utilities.datartorchmetrics.utilities.importsrtorchmetrics.utilities.plotrr__doctest_skip__rr.r.r.r6s