Unicode Decode Error with input audio file - preprocess module

[netolcc06]

Hi,

I'm trying to execute the demo.py according to your specifications, but I'm getting the following warning and errors from preprocess module:

preprocess: warn: defaulting content type for 'demo/demo.wav' to 'Text'

File "demo.py", line 70, in p.preprocess(infile, outfile) File "/.local/lib/python3.6/site-packages/preprocess.py", line 417, in preprocess lines = fin.readlines() File "/usr/lib/python3.6/codecs.py", line 321, in decode (result, consumed) = self._buffer_decode(data, self.errors, final) UnicodeDecodeError: 'utf-8' codec can't decode byte 0xaf in position 5: invalid start byte

Is the audio file supposed to be transformed or at least internally considered as text? Is this transformation related to the UnicodeDecodeError? @HsinYingLee

[netolcc06]

I've tried a few things and nothing worked, so I assume I'm not using the correct preprocess module. Could you share it with us?

[sunyasheng]

@netolcc06 Did you find their dataset? I did not find the dataset.

[xiaodongyang]

@sunyasheng dataset http://vllab.ucmerced.edu/hylee/Dancing2Music/script.txt

[ak9250]

@netolcc06 were you able to get the inference demo working?

@xiaodongyang is the preprocess module correct?

[xiaodongyang]

@ak9250 What errors did you get?

[ak9250]

@xiaodongyang see https://github.com/NVlabs/Dancing2Music/issues/1#issuecomment-626378974

[argentinox]

I have the same error! If I create a conda environment with python 2.7 I can install the correct preprocess module with "pip install preprocess"... after that I have other modules conflict problem

[HsinYingLee]

Hi all, The preprocess code is adopted from the Obama Libsync project. The proprocess.py and mfcc_rev.py can be found there.