ML-SAN uses input calibration with FiLM, interaction gating, and output regularization to adapt emotion recognition to individual speaker styles, reporting gains on MELD and IEMOCAP especially for rare sentiment classes.
In: ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ML-SAN: Multi-Level Speaker-Adaptive Network for Emotion Recognition in Conversations
ML-SAN uses input calibration with FiLM, interaction gating, and output regularization to adapt emotion recognition to individual speaker styles, reporting gains on MELD and IEMOCAP especially for rare sentiment classes.