Bayesian U-Net for Segmenting Glaciers in SAR Imagery

TL;DR

This paper introduces a Bayesian U-Net model that estimates uncertainty in glacier segmentation from SAR images, improving accuracy and aiding manual annotation processes.

Contribution

It presents a novel two-stage Bayesian approach using dropout in U-Net for uncertainty estimation in glacier segmentation.

Findings

Achieved 95.24% Dice similarity in segmentation.

Provided uncertainty maps to guide manual annotation.

Enhanced segmentation performance over deterministic U-Net.

Abstract

Fluctuations of the glacier calving front have an important influence over the ice flow of whole glacier systems. It is therefore important to precisely monitor the position of the calving front. However, the manual delineation of SAR images is a difficult, laborious and subjective task. Convolutional neural networks have previously shown promising results in automating the glacier segmentation in SAR images, making them desirable for further exploration of their possibilities. In this work, we propose to compute uncertainty and use it in an Uncertainty Optimization regime as a novel two-stage process. By using dropout as a random sampling layer in a U-Net architecture, we create a probabilistic Bayesian Neural Network. With several forward passes, we create a sampling distribution, which can estimate the model uncertainty for each pixel in the segmentation mask. The additional uncertainty map information can serve as a guideline for the experts in the manual annotation of the data. Furthermore, feeding the uncertainty map to the network leads to 95.24% Dice similarity, which is an overall improvement in the segmentation performance compared to the state-of-the-art deterministic U-Net-based glacier segmentation pipelines.