samplerate = None
Please leave your comments, it might help someone who encounters the same problem in the future.
Hallo mpariente,
I haven't seen the " enhance single" so I wanted to use the network for separation, and that didnt work obviously ;). The error occured that the samplerate would be of NoneType...
But if it is a network just for speech enhancement for only one source it makes sense, that separate() doesn't work.
But now that I found this pretrained network for speech enhancement, could you tell me the function I need to enhance a single source, for example the already separated one?
best regards MaReeentier
Hello mpariente;
I've tried some other "enhanced single" pretrained networks, and it worked with the function : separate(). But "mpariente/ConvTasNet_Libri1Mix_enhsingle_8k" still doesn't work. The following error comes up:
RuntimeError: Received a signal with a sampling rate of 16000Hz for a model of NoneHz. You can pass resample=True to resample automatically.
And if I pass resample=True to the separate function:
model.separate(mixture ,estimation, resample = True )
Still the error occures :
wav = _resample(wav[:, 0], orig_sr=fs, target_sr=int(model.sample_rate))[:, None]
_____________________ ^^^^^^^^^^^^^^^^^^^^^^
TypeError: int() argument must be a string, a bytes-like object or a real number, not 'NoneType'
So for me it looks like there is a problem with the type been set for the frequency rate in the model. Could that be possible? Or am I missing something?
Best regards
MaReeentier
Yes, there seems to be a problem, sorry.