4 ms·

Just wanted to mention there's some folks doing realtime source separation (not sure exactly how they've implemented it) with a DNN for reduction of background

Just wanted to mention there's some folks doing realtime source separation (not sure exactly how they've implemented it) with a DNN for reduction of background noise in, eg: Skype conversations. I'm not involved with them in any way, but I've been amazed with its ability to cancel out coffee-shop style noise. Check out https://krisp.ai/technology/ https://krisp.ai/technology/ - Mac/Windows. I wish they had Linux support! Edit: Appears they don't have Windows support yet.

brucemoose 8y ago
> Uninterrupted Voice The same krispNet DNN, trained on hundreds of hours of customized data, is able to perform Packet Loss Concealment (predicting lost network packets) for audio and fill out missing voice chunks by eliminating "chopping" in voice calls. This is both fascinating and horrifying at the same time! I wonder if/when it would be possible to rewrite whole words in real time using a voice that sounds just like you.