Fig. 22From: Intelligent radar HRRP target recognition based on CNN-BERT modelAverage attention map in the three encoder blocks of Cessna Citation S/II aircraft; a the shallow layer, b the middle layer, c the deep layerBack to article page