search

tyiannak / pyAudioAnalysis

Python Audio Analysis Library: Feature Extraction, Classification, Segmentation and Applications
Apache License 2.0
5.86k stars 1.19k forks source link

Speaker Diarization results #223

Open rakky18 opened 5 years ago

rakky18 commented 5 years ago

Hi, I am able to use the speaker diarization now and it is running good and produces results. But the results aren't accurate. I have an example wav file with only 1 speaker talking inside and when I run the speaker diarization, it got 6 speakers labelled. When I try to run again it got 5 speakers. It isn't accurate for me and I think I may be doing something wrong here? I didn't change anything on the codes though.

I hope someone would reply. I would be so grateful.

vnylp commented 5 years ago

@rakky18 hi can you help me in implementing this code?

giorgionanfa commented 5 years ago

I have the same problem...@rakky18

aakashak2000 commented 4 years ago

There's a hyperparameter within the speaker_diarization function called n_speakers. You need to mention the number of speakers speaking in the audio beforehand. in your case n_speakers should be 1. But generally this function is used for audio files with 2 or more speakers.