Trusted Fine-Grained Image Classification Based on Evidence Theory and Its Applications to Medical Image Analysis

Abstract

Fine-grained image classification (FGIC) aims to classify object of images to the corresponding subordinate classes of a superclass. Due to insufficient training data and confusing data samples, FGIC may produce uncertain classification results that are untrusted for data applications. Dempster-Shafer evidence theory (DST) is widely applied in reasoning with uncertainty and opinion fusion. Recently, researchers extended DST by combining it with deep learning to measure the uncertainty of deep neural networks and perform uncertainty classification. In this proposed chapter, the authors provide a detailed introduction of how to integrate ENN to construct the trusted FGIC model. Compared with the traditional approaches, the trusted FGIC method not only generates accurate classification results but also reduces the uncertainty of fine-grained classification. In addition, they introduce the application of FGIC in medical image analysis to achieve trusted fine-grained medical image classification.