Spatial-Channel Attention-Based Class Activation Mapping for Interpreting CNN-Based Image Classification Models

<div>ResNet-18 network: the feature map visualization of the highest convolutional layer (subfigure (b)). The class activation map visualization of GAP-CAM (c, d) and Grad-CAM (e, f). (a) Original image; (b) feature map visualization; (c) GAP-CAM:dog; (d) GAP-CAM:cat; (e) Grad-CAM:dog; (f) Grad-CAM:cat.</div>

Security and Communication Networks

fig2

Figure 2

Figure 2: Spatial-Channel Attention-Based Class Activation Mapping for Interpreting CNN-Based Image Classification Models 