Fig. 3From: Efficient emotion recognition using hyperdimensional computing with combinatorial channel encoding and cellular automataHDC a late fusion and b early fusion architectures for a three-modality emotion recognition system. The late fusion architecture fuses after the temporal encoder, resulting in 3 parallel temporal encoders—one per modality. In comparison, the early fusion architecture fuses before the temporal encoder, resulting in only 1 temporal encoderBack to article page