o %ݫi @sxdZddlZddlmZddlmZddlmZGdddej j Z Gddde Z Gd d d e Z Gd d d e ZdS) zb Combinations of processing algorithms to implement downsampling methods. Authors * Salah Zaiem N)Conv1d) Pooling1dc@seZdZdZddZdS) Downsamplerz#Wrapper for downsampling techniquescCs ||S)zDownsampling function Arguments --------- x : tensor Speech samples of shape [B,n_samples] with B the batch size Returns ------- Downsampled outputs. ) downsampler)selfxrR/home/ubuntu/.local/lib/python3.10/site-packages/speechbrain/lobes/downsampling.pyforwards zDownsampler.forwardN)__name__ __module__ __qualname____doc__r rrrr rs rc eZdZdZfddZZS)SignalDownsampleraSignal downsampling (Decimation) Arguments --------- downsampling_factor : int Factor of downsampling (i.e. ratio (length before ds / length after ds)) initial_sampling_rate : int Sampling_rate of the input audios Example ------- >>> sd = SignalDownsampler(2,16000) >>> a = torch.rand([8,28000]) >>> a = sd(a) >>> print(a.shape) torch.Size([8, 14000]) cs8t||_t|||_tj||jtjd|_ dS)N)dtype) super__init__downsampling_factorinttarget_ds_rateTResampletorchfloat32r)rrinitial_sampling_rate __class__rr r5s   zSignalDownsampler.__init__r r r rr __classcell__rrrr r"srcr)Conv1DDownsamplera1D Convolutional downsampling with a learned convolution Arguments --------- downsampling_factor : int Factor of downsampling (i.e. ratio (length before ds / length after ds)) kernel_size : int Kernel size of the 1D filter (must be an odd integer) Example ------- >>> sd = Conv1DDownsampler(3,161) >>> a = torch.rand([8,33000]) >>> a = sd(a) >>> print(a.shape) torch.Size([8, 10947]) cs6t||_||_t|jd|jdddgd|_dS)Nvalid)stridepadding kernel_size out_channels input_shape)rrr%rrr)rrr%rrr rPs  zConv1DDownsampler.__init__rrrrr r >sr cs$eZdZdZ dfdd ZZS)PoolingDownsamplera?1D Pooling downsampling (non-learned) Arguments --------- downsampling_factor : int Factor of downsampling (i.e. ratio (length before ds / length after ds)) kernel_size : int Kernel size of the 1D filter (must be an odd integer) padding : int The number of padding elements to apply. pool_type : string Pooling approach, must be within ["avg","max"] Example ------- >>> sd = PoolingDownsampler(3,41) >>> a = torch.rand([8,33000]) >>> a = sd(a) >>> print(a.shape) torch.Size([8, 10987]) ravgcsBt||_||_||_||_t|j|j|jd|jd|_dS)N)r#r$r% input_dims pool_type)rrr%r$r,rrr)rrr%r$r,rrr rss  zPoolingDownsampler.__init__)rr)rrrrr r(]sr()rrtorchaudio.transforms transformsrspeechbrain.nnet.CNNrspeechbrain.nnet.poolingrnnModulerrr r(rrrr s