Xianzhi Wang

DDE-SER: A Dual-Decomposition Ensemble Framework Fusing Adaptive Variational Modes and Harmonic-Percussive Spectrograms for Speech Emotion Recognition

: The accurate classification of human emotions from speech remains a formidable challenge due to the dynamic, non-stationary properties of audio signals and pervasive background noise. Traditional single-domain extraction methods frequently fail to capture overlapping acoustic phenomena, resulting in high misclassific...