o eif @s*dZddlZddlmZGdddZdS)zPreprocessors for audioN)Resamplec@s*eZdZdZd ddZddZdd Zd S) AudioNormalizeraUNormalizes audio into a standard format Arguments --------- sample_rate : int The sampling rate to which the incoming signals should be converted. mix : {"avg-to-mono", "keep"} "avg-to-mono" - add all channels together and normalize by number of channels. This also removes the channel dimension, resulting in [time] format tensor. "keep" - don't normalize channel information Example ------- >>> import torchaudio >>> example_file = 'tests/samples/multi-mic/speech_-0.82918_0.55279_-0.082918.flac' >>> signal, sr = torchaudio.load(example_file, channels_first = False) >>> normalizer = AudioNormalizer(sample_rate=8000) >>> normalized = normalizer(signal, sr) >>> signal.shape torch.Size([160000, 4]) >>> normalized.shape torch.Size([80000]) NOTE ---- This will also upsample audio. However, upsampling cannot produce meaningful information in the bandwidth which it adds. Generally models will not work well for upsampled data if they have not specifically been trained to do so. > avg-to-monocCs,||_|dvrtd|||_i|_dS)N)rkeepz Unexpected mixing configuration ) sample_rate ValueErrormix_cached_resamplers)selfrr r [/home/ubuntu/transcripts/venv/lib/python3.10/site-packages/speechbrain/dataio/preprocess.py__init__(s  zAudioNormalizer.__init__cCsD||jvrt||j|j|<|j|}||dd}||S)ayPerform normalization Arguments --------- audio : torch.Tensor The input waveform torch tensor. Assuming [time, channels], or [time]. sample_rate : int Rate the audio was sampled at. Returns ------- audio : torch.Tensor Channel- and sample-rate-normalized audio. r)r rr unsqueezesqueeze_mix)r audior resampler resampledr r r __call__/s   zAudioNormalizer.__call__cCs<|dk}|jdkr|r|St|dS|jdkr|SdS)zHandle channel mixingrrN)dimr torchmean)r r flat_inputr r r rHs    zAudioNormalizer._mixN)rr)__name__ __module__ __qualname____doc__rrrr r r r rs   r)rrspeechbrain.augment.time_domainrrr r r r s