3 ms·
Ivy Speech – Realtime audio processing to block or allow certain voices
- jimbooonooo 3y agoHowdy all, Recently, a friend and I were joking about ways to ignore unpleasant voices at a family BBQ. A quick look at current speech detection methods led down a rabbit-hole (of course), and resulted in this fun proof of concept. I was surprised at how well speaker embedding performed (and also acknowledge the sparse, subjective nature of my results). It was a fun weekend project, and I hope you enjoy the write up!