Hi , I’m currently using Deepgram STT with the nova-3 model and noticing some transcription issues. Sometimes the system hallucinates words (e.g., “book the demo” → “book the loan”), and when the audio volume is low it occasionally generates incorrect words. We also use domain-specific terms like demo, booking, appointment, meeting, and schedule, which sometimes get misrecognized. From the documentation I noticed features like keyterm prompting and endpointing (but it will cause latency right?)that might help. Are these the recommended approaches to reduce hallucination and improve transcription accuracy,low volume hallucination or are there other best practices for this scenario?
#Transcription Hallucination and incorrect word generates on low volume.
tame tapir
sharp steppeBOT
Thanks for asking your question. Please be sure to reply with as much detail as possible so the community can assist you efficiently.
-# If you haven't done so, ensure your Discord and Github profiles are linked to Deepgram so you can earn points to redeem on cool stuff just by being active!
Thanks for asking your question. Please be sure to reply with as much detail as possible so the community can assist you efficiently.
-# If you haven't done so, ensure your Discord and Github profiles are linked to Deepgram so you can earn points to redeem on cool stuff just by being active!
Hey there! You don't seem to have your Discord account connected to your Deepgram account. Not sure how you posted in here, but you should go to https://community.deepgram.com and link your accounts!
Hey there! You don't seem to have your Discord account connected to your Deepgram account. Not sure how you posted in here, but you should go to https://community.deepgram.com and link your accounts!