SmoothGrad: removing noise by adding noise

Code

• github.com/PAIR-code/saliency
  Officialtf★ 994
• github.com/shap/shap
  tf★ 25,171
• github.com/slundberg/shap
  tf★ 25,160
• github.com/pytorch/captum
  pytorch★ 5,583
• github.com/shaoshanglqy/shap-shapley
  tf★ 10
• github.com/miaolan-xie/shap
  tf★ 0
• github.com/austinbrown34/shap
  tf★ 0
• github.com/saivarunr/xshap
  tf★ 0
• github.com/idiap/fullgrad-saliency
  pytorch★ 0
• github.com/sicara/tf-explain
  tf★ 0

Abstract

Explaining the output of a deep network remains a challenge. In the case of an image classifier, one type of explanation is to identify pixels that strongly influence the final decision. A starting point for this strategy is the gradient of the class score function with respect to the input image. This gradient can be interpreted as a sensitivity map, and there are several techniques that elaborate on this basic idea. This paper makes two contributions: it introduces SmoothGrad, a simple method that can help visually sharpen gradient-based sensitivity maps, and it discusses lessons in the visualization of these maps. We publish the code for our experiments and a website with our results.

Tasks

Reproductions