Three-layered draw-attention model for humanoid robots with gestures and verbal cues | IEEE Conference Publication | IEEE Xplore