Learn about popular Explainable AI algorithms such as integrated gradient, LIME, class activation maps, counterfactual explanations, etc.

docker.tar.gz

Explainable AI is a set of tools and frameworks that helps you understand and interpret the internal logic behind the predictions made by a deep learning network. With this, you can generate insights into the behavior and working of the model to mitigate issues around it in the development phase.

In this course, you will be introduced to popular Explainable AI algorithms such as smooth gradient, integrated gradient, LIME, class activation maps, counterfactual explanations, feature attributions, etc., for image classification networks such as MobileNet-V2 trained on large-scale datasets like ImageNet-1K.

By the end of this course, you will understand the need for Explainable AI and be able to design and implement popular explanation algorithms like saliency maps, class activation maps, counterfactual explanations, etc. You will be able to evaluate and quantify the quality of the neural network explanations via several interpretability metrics.

Master Explainable AI: Interpreting Image Classifier Decisions

They suppress the gradients of negative activations and further backpropagation of negative gradients.

They reduce visual noise by averaging the vanilla gradients of multiple noisy versions of the input image. 

They capture the relevant gradients along a linear path joining the input image and a baseline (usually zero). 

They explain the predictions of any classifier by approximating it locally with an interpretable model.

Which statement is *true* about integrated gradients?

Which statement is true about integrated gradients?

**(Select all that apply.)** Which algorithms don't require gradient computation for generating a saliency map?

(Select all that apply.) Which algorithms don’t require gradient computation for generating a saliency map?

Which LRP rule favors positive contributions rather than negative ones in the saliency map?

Which LRP rule favors positive contributions rather than negative ones in the saliency map?

The relevances of all neurons in any layer should sum to the model prediction score.

The relevance of the output layer is set to zero for all neurons.

The relevance of neurons corresponding to the prediction in the output layer is the same as the model prediction score.

**(Select all that apply.)** Which statements are *true* about Layer-wise Relevance Propagation?

(Select all that apply.) Which statements are true about Layer-wise Relevance Propagation?

**(Select all that apply.)** What are the disadvantages of the vanilla gradient algorithm?

(Select all that apply.) What are the disadvantages of the vanilla gradient algorithm?

Quiz yourself on the concepts learned in this chapter.

Introduction to Explainable AI

Saliency Maps

Class Activation Maps

Miscellaneous Methods

Metrics of Interpretability

Final Assessment

Weakly Supervised Object Localization

Quiz on Saliency Maps