Why do we need Speech model for Speaker Identification?

alphacep / vosk-api

Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node

Apache License 2.0

7.7k stars 1.08k forks source link

Why do we need Speech model for Speaker Identification? #144

Closed omerasif-itu closed 4 years ago

omerasif-itu commented 4 years ago

In python/example/test_speaker.py, why do we need to provide a speech model with speaker model? As speaker identification is independent of speech model.

vosk-model-spk-0.3 | 13M | TBD | Model for speaker identification, should work for all languages

nshmyrev commented 4 years ago

Speech model is used to extract utterances (think of very advanced VAD).

omerasif-itu commented 4 years ago

Ahan. So I am assuming that one needs a Russian Model to identify Russian Speakers. Is it correct?

nshmyrev commented 4 years ago

You can probably use any vad, we just don't have this in the API.

omerasif-itu commented 4 years ago

Alright. Thanks!