method · first seen 9 Oct 2026
ASR confidence filtering
Selecting unlabeled training audio by automatic speech recognition confidence as a quality proxy.
Mentions per week
PapersCodeDiscussion
Earliest sources
- From 2.1M Hours of Raw Broadcast Audio to Finnish Speech SSL: Confidence Filtering and Architectural Analysis at ScalePapers · 9 Oct 2026 · Yaroslav Getman, Tamás Grósz, Tommi Lehtonen, Mikko Kurimo
Latest sources
- From 2.1M Hours of Raw Broadcast Audio to Finnish Speech SSL: Confidence Filtering and Architectural Analysis at ScalePapers · 9 Oct 2026 · Yaroslav Getman, Tamás Grósz, Tommi Lehtonen, Mikko Kurimo