CALM jointly models acoustic speaker identity and linguistic context to cut biased error rates by more than half on two-speaker English and Japanese mixtures.
Frame-level target- speaker activity posteriors are computed as: P vad =σ(W vad ˆH(L) +b vad),(11) withP vad ∈[0,1] T enc
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CALM: Joint Contextual Acoustic-Linguistic Modeling for Personalization of Multi-Speaker ASR
CALM jointly models acoustic speaker identity and linguistic context to cut biased error rates by more than half on two-speaker English and Japanese mixtures.