Implementations

Included with this toolkit are a number of implementations for the interfaces described in the previous section. Unlike the interfaces, which declare operation and use case, implementations provide variations on how to satisfy the interface-defined use case, varying trade-offs, or results implications.

Image Perturbation

`xaitk_saliency.impls.perturb_image.mc_rise`	Implementation of MC-RISE perturbation mask generation
`xaitk_saliency.impls.perturb_image.random_grid`
`xaitk_saliency.impls.perturb_image.rise`	RISE MIT License (https://github.com/eclique/RISE/blob/master/LICENSE):
`xaitk_saliency.impls.perturb_image.sliding_radial`	This module defines the SlidingRadial class, which implements a sliding radial perturbation on input images.
`xaitk_saliency.impls.perturb_image.sliding_window`	This module defines the SlidingWindow class, which implements a sliding window perturbation on input images.

Heatmap Generation

`xaitk_saliency.impls.gen_detector_prop_sal.drise_scoring`	Implementation of DRISE scorer
`xaitk_saliency.impls.gen_classifier_conf_sal.mc_rise_scoring`	Implementation of MC-RISE scorer
`xaitk_saliency.impls.gen_classifier_conf_sal.occlusion_scoring`	Implementation of OcclusionScoring scorer
`xaitk_saliency.impls.gen_classifier_conf_sal.rise_scoring`	Implementation of RISEScoring scorer
`xaitk_saliency.impls.gen_descriptor_sim_sal.similarity_scoring`	Implementation of SimilarityScoring scorer
`xaitk_saliency.impls.gen_classifier_conf_sal.squared_difference_scoring`	Implementation of SimilarityScoring scorer

End-to-End Saliency Generation

Image Classification

`xaitk_saliency.impls.gen_image_classifier_blackbox_sal.mc_rise`	Implementation of MC-RISE saliency stack
`xaitk_saliency.impls.gen_image_classifier_blackbox_sal.occlusion_based`	This module defines the PerturbationOcclusion class, which implements a generator composed of modular perturbation and occlusion-based algorithms
`xaitk_saliency.impls.gen_image_classifier_blackbox_sal.rise`	Encapsulation of the perturbation-occlusion method using specifically the RISE implementations of the component algorithms.
`xaitk_saliency.impls.gen_image_classifier_blackbox_sal.slidingwindow`	Encapsulation of the perturbation-occlusion method using specifically sliding windows and the occlusion-scoring method.

Image Similarity

`xaitk_saliency.impls.gen_image_similarity_blackbox_sal.occlusion_based`	This module defines the PerturbationOcclusion class, which implements a generator composed of modular perturbation and occlusion-based algorithms
`xaitk_saliency.impls.gen_image_similarity_blackbox_sal.sbsm`	This module defines SBSMStack, which implements the perturbation-occlusion method using specifically the sliding window image perturbation and similarity scoring algorithms to generate similarity-based visual saliency maps

Object Detection

`xaitk_saliency.impls.gen_object_detector_blackbox_sal.occlusion_based`	This module defines the PerturbationOcclusion class, which implements a generator composed of modular perturbation and occlusion-based algorithms
`xaitk_saliency.impls.gen_object_detector_blackbox_sal.drise`	This module defines _BaseDRISE, DRISEStack, and RandomGridStack.

Saliency Metric

xaitk_saliency.impls.saliency_metric.entropy

Provides an implementation of the SaliencyMetric interface for computing entropy over a saliency map using scipy.stats.entropy.