winston-lin-wei-cheng / opensmile-llds-extraction Goto Github PK

I am using the following code, assuming my wav files are stored in the RENAMED_wav folder, which is actually in that location. However, where are you storing the labels for corresponding files?

  opensmile_root = '...../opensmile/build/progsrc/smilextract'
  input_audio_folder_path = ........./RENAMED_wav/'
  feature_type ='IS13ComParE'               # 'IS13ComParE' or 'eGeMAPS' are currently supported 
  
  # create output directory and folders
  os.mkdir(input_audio_folder_path.replace('wav',feature_type+'_llds'))
  os.mkdir(input_audio_folder_path.replace('wav',feature_type+'_llds')+'/feat_arff/')
  os.mkdir(input_audio_folder_path.replace('wav',feature_type+'_llds')+'/feat_mat/')

The code is successfully the folders feat_arff and feat_mat But they are empty,
I can see the openSmile is extracting 103 components in the console.
What is the error?

Please guide

Hi

I have been working with the IS13_ComParE.conf and eGeMAPSv01a.conf files from the "/config/is09-13/" and "/config/egemaps/v01a" directories, respectively.

In my attempts to use these configuration files, I encountered the following error: "data = arff.loadarff(open ("temp.arff"))
File "/root/.local/share/virtualenvs/workspace-hArTflJD/lib/python3.10/site-packages/scipy/io/arff/_arffread.py", line 802, in loadarff
return _loadarff(ofile)
File "/root/.local/share/virtualenvs/workspace-hArTflJD/lib/python3.10/site-packages/scipy/io/arff/_arffread.py", line 835, in _loadarff
raise NotImplementedError("String attributes not supported yet, sorry")
NotImplementedError: String attributes not supported yet, sorry"

I noticed that you have a "config_lld" folder in your repository. Could you please let me know where I can find this "config_lld" folder? I am curious if there are specific configurations or adjustments in that folder that might resolve the error I'm encountering.