search

group-24 / Palpitate

45 stars 16 forks source link

This is the script I started to write in python. For now it just take… #18

Open mm8112 opened 9 years ago

mm8112 commented 9 years ago

…s a list of (SubjectID,Path_to_audio_file) pairs (from the terminal) and calls Opensmile for each file, using the MBC config file and outputting and .arff file with all the attributes of the audio file. Before I can change the script, I need to change some features in the MBC config file, particularly concerning the frame size

kren1 commented 9 years ago

Awsome, how does this deal with time slicing? I see there is no need to convert these files to weka, which is good news!

mm8112 commented 9 years ago

Yeah, the format of the output file is perfect. Before I start dealing with timeslices, I need to change the MBC config file on opensmile. At the moment the file does not even cut the audio file into slices, it just analyses it as a whole (because some features like framesize are set to 0). After changing the config file to include timeslices I will add more stuff to the script

kren1 commented 9 years ago

Can you also add the config file to the commit? and move both files in a separate folder?