Fig. 7From: Intelligent radar HRRP target recognition based on CNN-BERT modelOutput sequence features a map of convolutional token embedding module; left: two-dimensional feature map, right: single channel featureBack to article page