The CNN mainly extracts feature regions of interest and Transformer encoder is utilized to assign high weights to the regions of interest. The CNN-Transformer model also includes the novel S-shaped ...