Contributo in atti di convegno, 2021, ENG, 10.1109/ISCC53001.2021.9631485

Exemplars and counterexemplars explanations for image classifiers, targeting skin lesion labeling

Metta C.; Guidotti R.; Yin Y.; Gallinari P.; Rinzivillo S.

CNR-ISTI, Pisa, Italy; University of Pisa, Pisa, Italy; Sorbonne Universite, Paris, France; Sorbonne Universite and Criteo AI Lab, Paris, France; CNR-ISTI, Pisa, Italy

Explainable AI consists in developing mechanisms allowing for an interaction between decision systems and humans by making the decisions of the formers understandable. This is particularly important in sensitive contexts like in the medical domain. We propose a use case study, for skin lesion diagnosis, illustrating how it is possible to provide the practitioner with explanations on the decisions of a state of the art deep neural network classifier trained to characterize skin lesions from examples. Our framework consists of a trained classifier onto which an explanation module operates. The latter is able to offer the practitioner exemplars and counterexemplars for the classification diagnosis thus allowing the physician to interact with the automatic diagnosis system. The exemplars are generated via an adversarial autoencoder. We illustrate the behavior of the system on representative examples.

ISCC 2021 - IEEE Symposium on Computers and Communications, Athens, Greece, 5-8/09/2021

Keywords

Image classification, Explainable AI, Machine Learning, Skin lesion image classification

CNR authors

Metta Carlo, Rinzivillo Salvatore

CNR institutes

ISTI – Istituto di scienza e tecnologie dell'informazione "Alessandro Faedo"

ID: 464865

Year: 2021

Type: Contributo in atti di convegno

Creation: 2022-03-06 22:10:48.000

Last update: 2024-01-30 15:12:47.000

External IDs

CNR OAI-PMH: oai:it.cnr:prodotti:464865

DOI: 10.1109/ISCC53001.2021.9631485

Scopus: 2-s2.0-85123215030

ISI Web of Science (WOS): 000936276000113