Techniques for Convolutional Neural Network Interpretability
Saliency maps for Convolutional Neural Networks are a technique used to visualize and understand the behavior of CNN models, particularly in tasks such as image classification, object detection and segmentation. They highlight which parts of an ...
