Interpretable Saliency Maps And Self-Supervised Learning For Generalized Zero Shot Medical Image Classification

TL;DR

This paper introduces a generalized zero-shot learning approach for medical image classification that leverages self-supervised learning to identify class representatives and generate synthetic features, eliminating the need for class attribute vectors.

Contribution

It presents a novel SSL-based GZSL method for medical images that outperforms existing methods and does not require class attribute vectors, unlike prior approaches.

Findings

Matches state-of-the-art SSL GZSL for natural images

Outperforms all existing methods for medical images

Uses a simpler architecture

Abstract

In many real world medical image classification settings we do not have access to samples of all possible disease classes, while a robust system is expected to give high performance in recognizing novel test data. We propose a generalized zero shot learning (GZSL) method that uses self supervised learning (SSL) for: 1) selecting anchor vectors of different disease classes; and 2) training a feature generator. Our approach does not require class attribute vectors which are available for natural images but not for medical images. SSL ensures that the anchor vectors are representative of each class. SSL is also used to generate synthetic features of unseen classes. Using a simpler architecture, our method matches a state of the art SSL based GZSL method for natural images and outperforms all methods for medical images. Our method is adaptable enough to accommodate class attribute vectors when they are available for natural images.