A new semi-supervised ASR framework, TopIPL, combines a dynamic pseudo-label cache and top-N checkpoint teacher averaging to improve WER by up to 40 percent in low-resource settings.
Semi-supervised learning miti- gates this challenge by leveraging large-scale unlabeled speech data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Unified Semi-Supervised Pipeline for Automatic Speech Recognition
A new semi-supervised ASR framework, TopIPL, combines a dynamic pseudo-label cache and top-N checkpoint teacher averaging to improve WER by up to 40 percent in low-resource settings.