Closed vmazashvili closed 3 months ago
let generate_saliency_map()
return the classes too. this way, I can grab these classes for the class_idx in saliency_map = cam(class_idx=[sample_class], scores=output)
for the new function generate_saliency_map_single()
call
plot_saliency_and_mask
after finetuning so that I can output the activation map for the masked image too