search

alphacep / vosk-api

Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node
Apache License 2.0
8k stars 1.11k forks source link

Ability to configure silence duration for endpoint detection #829

Closed muety closed 2 years ago

muety commented 2 years ago

Does Vosk server provide a way to configure the threshold for silence in seconds after which a new sentence is started?

nshmyrev commented 2 years ago

You can edit silence thresholds in model/model.conf

For details you can read https://kaldi-asr.org/doc/structkaldi_1_1OnlineEndpointRule.html

nshmyrev commented 2 years ago

Same as #380