echo cancellation while transcribing
Hey, I am trying to make an app that records system audio and mic audio on macOS and then transcribes them both separately.
The problem is that if I play a YouTube video out loud, the system audio leaks into the mic audio, and WhisperKit transcribes it.
I have tried several echo cancellation solutions, but they are not really working. Is there a way for WhisperKit to ignore background noise/audio and only transcribe my voice?
This is a very interesting advanced use case, thanks for the description @kanishkave. Audio denoising (from background speech and other noise) is on our roadmap but not yet scoped so we can not share an ETA yet. Do keep us posted in case you find a short-term solution for this.
hi @kanishkave , have you tried the default echo cancellation offered by macOS voice processing?
Hi @kanishkave were you able to figure out how to fix this issue? I'm also having issues with AEC.