Speech emotion recognition in real static and dynamic human-robot interaction scenarios

Nicolás Grágeda, Carlos Busso, Eduardo Alvarado, Ricardo García, Rodrigo Mahú, Fernando Huenupán, Néstor Becerra Yoma

Abstract: Highlights•Speech emotion recognition in real static and dynamic HRI scenarios was addressed.•Two classifiers were evaluated: ladder network and wav2vec.•Training conditions considered the acoustic model of the target room and beamforming responses.•Simulating the acoustic environment in the training dataset provided the highest average CCC.