Would it be able to feed in the audio itself for classifier_skid.compute_similarity function instead of giving the audio's path?

wenet-e2e / wespeaker

Research and Production Oriented Speaker Verification, Recognition and Diarization Toolkit

Apache License 2.0

664 stars 112 forks source link

Open DeliJingyiC opened 2 months ago

DeliJingyiC commented 2 months ago

classifier_skid = wespeaker.load_model('english')

JiJiJiang commented 2 months ago

Of course it can, welcome to contribute! The audio can be either numpy array or torch tensor.