Enable speaker labels and control how profanity and filler words are handled
We have introduced a More Option field in Services that allows users to enable speaker diarization, as well as control how profanity and filler words are handled. Not all engines support these options and some are enabled by default when supported. You have the ability to enable and disable each of these options by expanding the "More Options" section and selecting the corresponding option in the Transcript section of a service.
Diarization - Break up dialogue by speaker using speaker change indicators or chevrons.
Filter Profanity - Profanities are identified and filtered out from output
Filter Filler Words - Filler words, like um, uh, ah are identified and filtered out from output.
IMPORTANT: The Profanity filter is only available for the Speechmatics engine when using English (general), Spanish or Italian as the source language. The Filler Words filter is only available when using English (general) as the source language and either Speechmatics or Deepgram as the speech engine.
IMPORTANT: Diarization feature is only available for the Speechmatics engine.