MDAVIF: A Multi-Domain Acoustical-Visual Information Fusion Model for Depression Recognition from Vlog Data | IEEE Conference Publication | IEEE Xplore