Table 1.

Summary of explainable AI techniques classified according to Section 3.

Explanation Type	Paper	Technique	Intrinsic	Post Hoc	Global	Local	Model-Specify	Model-Agnostic
Feature	[28]	BP		*		*	*
	[29]	Guided-BP		*		*	*
	[30]	Deconv Network		*		*	*
	[31]	LRP		*		*	*
	[32]	CAM		*		*	*
	[33]	Grad-CAM		*		*	*
	[34]	LIME		*		*		*
	[35]	GraphLIME		*		*		*
	[36]	SHAP		*		*		*
	[37]	Attention	*			*	*
Example-based	[38]	ProtoPNet	*			*	*
	[39]	Triplet Network	*		*	*	*
	[5]	xDNN	*			*	*
Textual	[40]	TCAV		*	*	*		*
	[41]	Image Captioning	*			*	*

“*” indicates it belongs to this category, which is defined in Section 3, BP: backpropagation, CAM: class activation map, LRP: layer-wise relevance propagation, LIME: local interpretable model-agnostic explanations, MuSE: model usability evaluation, SHAP: Shapley additive explanations, xDNN: explainable deep neural network, TCAV: testing with concept activation vectors.