Deep Bayesian Active-Learning-to-Rank for Endoscopic Image Data

TL;DR

This paper introduces a deep Bayesian active-learning-to-rank framework that efficiently trains neural networks for endoscopic image severity estimation using relative annotations, especially effective with class imbalance.

Contribution

It proposes a novel active learning method that automatically selects informative image pairs for relative annotation in a ranking framework, improving efficiency and handling class imbalance.

Findings

Effective in endoscopic ulcerative colitis image analysis

Handles severe class imbalance well

Reduces annotation effort through active pair selection

Abstract

Automatic image-based disease severity estimation generally uses discrete (i.e., quantized) severity labels. Annotating discrete labels is often difficult due to the images with ambiguous severity. An easier alternative is to use relative annotation, which compares the severity level between image pairs. By using a learning-to-rank framework with relative annotation, we can train a neural network that estimates rank scores that are relative to severity levels. However, the relative annotation for all possible pairs is prohibitive, and therefore, appropriate sample pair selection is mandatory. This paper proposes a deep Bayesian active-learning-to-rank, which trains a Bayesian convolutional neural network while automatically selecting appropriate pairs for relative annotation. We confirmed the efficiency of the proposed method through experiments on endoscopic images of ulcerative colitis. In addition, we confirmed that our method is useful even with the severe class imbalance because of its ability to select samples from minor classes automatically.