speechmatics-python
speechmatics-python copied to clipboard
DEL-24775 - Support Multichannel Mod
This change introduces support for multichannel mode. Behaviour is explained in https://speechmatics.atlassian.net/wiki/spaces/~712020b3713ebb71814ed991afdd6d43fc1ac6/pages/4717805569/Summary+of+changes+-+Realtime+Multichannel+and+Channel+Diarization
This current implementation allows users to specify multichannel mode using the flag and arguments --multichannel <channel_name_1>,<channel_name_2>...<channel_name_n>. This must also include a --diarization type of either channelor channel_and_speaker.
As the engine configures all recognisers for multichannel to use the same audio format, audio formats of all inputs must also be matched with each other.