little help would be appreciated

[sangfrois]

Hey @kaplanan ,

I find your model very interesting, though, it seems like there are some parts missing. So if I want to access your model, I have to:

first get both databases

second download openSMILE and extract features from .wav files to .arff with the .conf files in your openSMILE directory (here I only see recola)

third write .npy files with .arff output by your script in "preprocessing" directory

only then could I have the fully trained model, i.e. the model parameters?

Thanks for your precisions on this!