Close
  Indian J Med Microbiol
 

Figure 6: Results of DeepCIN. From top to bottom, each column presents original image, localized vertical regions, contribution of segments within an image toward the image-level CIN classification. (represented as probability distribution over the segments. [attentional weights], the dotted lines indicate mean value and segments above the mean value, highlighted in green, are contributing the most), and corresponding ground truth and prediction labels, respectively

Figure 6: Results of DeepCIN. From top to bottom, each column presents original image, localized vertical regions, contribution of segments within an image toward the image-level CIN classification. (represented as probability distribution over the segments. [attentional weights], the dotted lines indicate mean value and segments above the mean value, highlighted in green, are contributing the most), and corresponding ground truth and prediction labels, respectively