Pseudo-labeling plus coarse label-aware contrastive pretraining on SAP and LibriSpeech yields a Whisper-based DSQA model with average SRCC 0.761 on five unseen multi-etiology, multi-language test sets.
Rank-N-Contrast: Learning continuous representations for regression
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Something from Nothing: Data Augmentation for Robust Severity Level Estimation of Dysarthric Speech
Pseudo-labeling plus coarse label-aware contrastive pretraining on SAP and LibriSpeech yields a Whisper-based DSQA model with average SRCC 0.761 on five unseen multi-etiology, multi-language test sets.