need help for a  use-case example

Thanks for the job.  we have a use-case that to judge an audio file that if it contains more then one speaker,  so that we can select out only one-speaker samples.  I do not know if this project can be a resolution. if it is ,would you please give us a inference example ? we need only a boolean result, True or False, and do not care which time step the speaker changed. thank you