I wonder how the training works in that system | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		cblconfederate on Aug 15, 2021 \| parent \| context \| favorite \| on: SoundStream: An End-to-End Neural Audio Codec I wonder how the training works in that system

rcthompson on Aug 15, 2021 [–]

I assume think they start with the "signal" and "noise" as separate audio files, and then they play them together in order to create a synthetic noisy input. Then they can train the output against only the signal so that it will learn to filter out the noise.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact