search

Atul-Anand-Jha / Speaker-Identification-Python

Speaker Identification System (upto 100% accuracy); built using Python 2.7 and python_speech_features library
GNU Lesser General Public License v3.0
206 stars 75 forks source link

Speaker identification feature? #26

Closed heedpraveen closed 4 years ago

heedpraveen commented 4 years ago

MFCC and Delta MFCC are the only features we need to focus on while building model on speaker identification?

If I use any different voice other than trained one it is showing the labels that is in training data and it is wrong, could you let me know why? and what to do if I need to get the model output as unknown if tested as stated?

Atul-Anand-Jha commented 4 years ago

You just have to set a threshold for the cosine difference value. Below that, you can return "UNKNOWN" as the identified speaker.

And, MFCC and Delta MFCC is an example for set of features for the voice data.Definitely, there are many othe