Opti-CAM: Optimizing saliency maps for interpretability

arXiv:2301.07002v2 Announce Type: replace
Abstract: Methods based on class activation maps (CAM) provide a simple mechanism to interpret predictions of convolutional neural networks by using linear combinations of feature maps as saliency maps. By contrast, masking-based methods optimize a saliency map directly in the image space or learn it by training another network on additional data.
In this work we introduce Opti-CAM, combining ideas from CAM-based and masking-based approaches. Our saliency map is a linear combination of feature maps, where …

abstract arxiv class contrast convolutional neural networks cs.cv feature image interpretability learn linear map maps masking network networks neural networks predictions simple space training type