0
$\begingroup$

I would like to find a way to isolate the speech of each of the people in an audio record so I can create a file of that form :

[
   {
       "voice_fingerprint": "701066EDD3A0A40A2F53F61EAFD0E6AB",
       "sentences": {
           {
               "sentence": "do you like red apples",
               "position": 1.39 // Seconds. Time position in the audio record
           },
           {
               "sentence": "and how do you feel about time shifts",
               "position": 7.21
           }
       }
   },
   {
       "voice_fingerprint": "8FFEA051AF3E3FB9A80A51A98FE05896",
       "sentences": {
           {
               "sentence": "yes I do like them",
               "position": 4.54
           },
           {
               "sentence": "i feel well about traveling",
               "position": 10.18
           }
       }
   }
]

This may be an interview record.

The problem IS NOT the Speech to Text, but to isolate the two people's sentences. Preferably in Python.

Have you ever worked on this ? Do you have any hints ?

$\endgroup$
  • $\begingroup$ For anyone working on this kind of topic : the technology is called "diarisation". But this is quite new and even Google or IBM APIs don't give good results. $\endgroup$ – secavfr Oct 25 '19 at 22:51
0
$\begingroup$

The task of isolating 2 or more speakers is called speaker diarization, here a list of softwares and useful resources.

Once you have the 2 or more audio files containing the individual voices, you could run some speech-to-text network that also outputs time information.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.