Hyper Evidential Deep Learning to Quantify Composite Classification Uncertainty
This work addresses uncertainty quantification for composite classification in deep learning, which is important for applications like medical imaging or autonomous systems where ambiguous classes exist, but it is incremental as it builds on existing evidential deep learning and belief theory frameworks.
The paper tackles the problem of quantifying uncertainty in deep neural networks when classes have similar visual features, requiring composite labels, by proposing a Hyper-Evidential Neural Network that models predictive uncertainty using Subjective Logic and introduces vagueness as a new uncertainty type, demonstrating that it outperforms state-of-the-art methods on four image datasets.
Deep neural networks (DNNs) have been shown to perform well on exclusive, multi-class classification tasks. However, when different classes have similar visual features, it becomes challenging for human annotators to differentiate them. This scenario necessitates the use of composite class labels. In this paper, we propose a novel framework called Hyper-Evidential Neural Network (HENN) that explicitly models predictive uncertainty due to composite class labels in training data in the context of the belief theory called Subjective Logic (SL). By placing a grouped Dirichlet distribution on the class probabilities, we treat predictions of a neural network as parameters of hyper-subjective opinions and learn the network that collects both single and composite evidence leading to these hyper-opinions by a deterministic DNN from data. We introduce a new uncertainty type called vagueness originally designed for hyper-opinions in SL to quantify composite classification uncertainty for DNNs. Our results demonstrate that HENN outperforms its state-of-the-art counterparts based on four image datasets. The code and datasets are available at: https://github.com/Hugo101/HyperEvidentialNN.