Combining sustained vowel and sentence reading recordings through intermediate feature fusion with transformers improves voice pathology detection AUC by up to 13% over single-source methods.
1555 11/10/2022, PE00000013), is partially supported by the ”D.A.R.E
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
-
MVP: Multi-source Voice Pathology detection
Combining sustained vowel and sentence reading recordings through intermediate feature fusion with transformers improves voice pathology detection AUC by up to 13% over single-source methods.