o .wi@sddlmZddlmZmZmZddlmZmZddl m Z ddl m Z ddl mZmZddlmZmZdd giZes>d gZGd dde Zd S) )Sequence)AnyOptionalUnion)Tensortensor)$perceptual_evaluation_speech_quality)Metric)_MATPLOTLIB_AVAILABLE_PESQ_AVAILABLE)_AX_TYPE_PLOT_OUT_TYPE!PerceptualEvaluationSpeechQualitypesq&PerceptualEvaluationSpeechQuality.plotc seZdZUdZeed<eed<dZeed<dZeed<dZ eed<d Z e ed <d Z e ed < d de dede deddf fdd ZdededdfddZdefddZd!deeeedfdeedefddZZS)"ra Calculate `Perceptual Evaluation of Speech Quality`_ (PESQ). It's a recognized industry standard for audio quality that takes into considerations characteristics such as: audio sharpness, call volume, background noise, clipping, audio interference etc. PESQ returns a score between -0.5 and 4.5 with the higher scores indicating a better quality. This metric is a wrapper for the `pesq package`_. Note that input will be moved to ``cpu`` to perform the metric calculation. As input to ``forward`` and ``update`` the metric accepts the following input - ``preds`` (:class:`~torch.Tensor`): float tensor with shape ``(...,time)`` - ``target`` (:class:`~torch.Tensor`): float tensor with shape ``(...,time)`` As output of `forward` and `compute` the metric returns the following output - ``pesq`` (:class:`~torch.Tensor`): float tensor of PESQ value reduced across the batch .. hint:: Using this metrics requires you to have ``pesq`` install. Either install as ``pip install torchmetrics[audio]`` or ``pip install pesq``. ``pesq`` will compile with your currently installed version of numpy, meaning that if you upgrade numpy at some point in the future you will most likely have to reinstall ``pesq``. .. caution:: The ``forward`` and ``compute`` methods in this class return a single (reduced) PESQ value for a batch. To obtain a PESQ value for each sample, you may use the functional counterpart in :func:`~torchmetrics.functional.audio.pesq.perceptual_evaluation_speech_quality`. Args: fs: sampling frequency, should be 16000 or 8000 (Hz) mode: ``'wb'`` (wide-band) or ``'nb'`` (narrow-band) keep_same_device: whether to move the pesq value to the device of preds n_processes: integer specifying the number of processes to run in parallel for the metric calculation. Only applies to batches of data and if ``multiprocessing`` package is installed. kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info. Raises: ModuleNotFoundError: If ``pesq`` package is not installed ValueError: If ``fs`` is not either ``8000`` or ``16000`` ValueError: If ``mode`` is not either ``"wb"`` or ``"nb"`` Example: >>> from torch import randn >>> from torchmetrics.audio import PerceptualEvaluationSpeechQuality >>> preds = randn(8000) >>> target = randn(8000) >>> pesq = PerceptualEvaluationSpeechQuality(8000, 'nb') >>> pesq(preds, target) tensor(2.2885) >>> wb_pesq = PerceptualEvaluationSpeechQuality(16000, 'wb') >>> wb_pesq(preds, target) tensor(1.6805) sum_pesqtotalFfull_state_updateis_differentiableThigher_is_bettergplot_lower_boundg@plot_upper_boundfsmode n_processeskwargsreturnNc stjd i|tstd|dvrtd|||_|dvr(td|||_t|ts;|dkr;td|||_ |j dt d d d |j d t dd d dS)NzPerceptualEvaluationSpeechQuality metric requires that `pesq` is installed. Either install as `pip install torchmetrics[audio]` or `pip install pesq`.)i@i>z:Expected argument `fs` to either be 8000 or 16000 but got )wbnbz;Expected argument `mode` to either be 'wb' or 'nb' but got rzCExpected argument `n_processes` to be an int larger than 0 but got rgsum)defaultdist_reduce_fxr) super__init__r ModuleNotFoundError ValueErrorrr isinstanceintr add_stater)selfrrrr __class__r#T/home/ubuntu/sommelier/.venv/lib/python3.10/site-packages/torchmetrics/audio/pesq.pyr%bs z*PerceptualEvaluationSpeechQuality.__init__predstargetcCsJt|||j|jd|j|jj}|j|7_|j| 7_dS)z*Update state with predictions and targets.FN) rrrrtordevicer rnumel)r+r/r0 pesq_batchr#r#r.update|s z(PerceptualEvaluationSpeechQuality.updatecCs |j|jS)zCompute metric.)rr)r+r#r#r.computes z)PerceptualEvaluationSpeechQuality.computevalaxcCs |||S)abPlot 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.audio import PerceptualEvaluationSpeechQuality >>> metric = PerceptualEvaluationSpeechQuality(8000, 'nb') >>> metric.update(torch.rand(8000), torch.rand(8000)) >>> fig_, ax_ = metric.plot() .. plot:: :scale: 75 >>> # Example plotting multiple values >>> import torch >>> from torchmetrics.audio import PerceptualEvaluationSpeechQuality >>> metric = PerceptualEvaluationSpeechQuality(8000, 'nb') >>> values = [ ] >>> for _ in range(10): ... values.append(metric(torch.rand(8000), torch.rand(8000))) >>> fig_, ax_ = metric.plot(values) )_plot)r+r7r8r#r#r.plots &r)r)NN)__name__ __module__ __qualname____doc__r__annotations__rboolrrrfloatrr)strrr%r5r6rrrr r r: __classcell__r#r#r,r.rs0 ;      2N)collections.abcrtypingrrrtorchrr"torchmetrics.functional.audio.pesqrtorchmetrics.metricr torchmetrics.utilities.importsr r torchmetrics.utilities.plotr r __doctest_requires____doctest_skip__rr#r#r#r.s