Currently I am working with two large speech databases and I was asked to build one subset from each one in order to get two representative subsets with the same number of utterances. My question is: the number of utterances is the same as the number of audio files?

Thank you so much!

  • 1
    $\begingroup$ I can't answer that without knowing more about the databases in question. Is there no documentation which describes the make-up of the audio files? $\endgroup$ – Oliver Mason Apr 27 '18 at 15:57
  • $\begingroup$ @OliverMason ,Indeed! $\endgroup$ – quintumnia Apr 27 '18 at 16:48
  • 2
    $\begingroup$ Technically, this is a question about the file structure, as opposed to the AI algorithm, so Computer Science or Overflow may be more suitable. (Make sure to include the 3rd party software you are using.) Nevertheless, this has been upvoted, so I'll leave it up for the time being. But you do need to provided the details requested in the previous comments. $\endgroup$ – DukeZhou Apr 27 '18 at 20:51
  • $\begingroup$ To add detail to @OliverMason’s comment, an utterance would typically be one statement or sentence - does that align with one audio file? $\endgroup$ – G__ Apr 28 '18 at 5:12
  • $\begingroup$ The databases are WSJ and Globalphone for German and Spanish. Do you know something about them? $\endgroup$ – little_mice Apr 30 '18 at 10:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.