Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2010.12096/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTImproving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data — mentioned by nshmhttps://news.ycombinator.com/item?id=24898195urn:hackerpapers:mention:2010.12096:24898195Mon, 26 Oct 2020 16:48:55 GMT<p>Meanwhile Google trains their systems on 3 million hours of Youtube audio gaining 30% better accuracy: https://arxiv.org/abs/2010.12096 It improves much faster than you might think.</p><p><a href="https://hackerpaper.kaipfister.de/p/2010.12096/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.SD