ICA and whitening preprocessing before k-means improves DSU-based ASR for XLS-R-300M, and ICA axes show interpretable phonetic contrasts.
ML-SUPERB: Multilingual speech universal performance benchmark,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
eess.AS 1years
2025 1verdicts
CONDITIONAL 1roles
dataset 1polarities
use dataset 1representative citing papers
citing papers explorer
-
Discrete Speech Unit Extraction via Independent Component Analysis
ICA and whitening preprocessing before k-means improves DSU-based ASR for XLS-R-300M, and ICA axes show interpretable phonetic contrasts.