The Weakly-Labeled Rand Index

TL;DR

This paper introduces a new weakly-labeled Rand index for evaluating segmentation quality in SAS imagery with uncertain regions, outperforming traditional metrics in such contexts.

Contribution

A novel weakly-labeled Rand index and labeling approach tailored for SAS imagery with uncertain regions, improving segmentation evaluation accuracy.

Findings

The new index aligns well with qualitative segmentation assessments.

It outperforms traditional metrics on weakly-labeled SAS data.

The approach effectively handles regions of uncertainty in imagery.

Abstract

Synthetic Aperture Sonar (SAS) surveys produce imagery with large regions of transition between seabed types. Due to these regions, it is difficult to label and segment the imagery and, furthermore, challenging to score the image segmentations appropriately. While there are many approaches to quantify performance in standard crisp segmentation schemes, drawing hard boundaries in remote sensing imagery where gradients and regions of uncertainty exist is inappropriate. These cases warrant weak labels and an associated appropriate scoring approach. In this paper, a labeling approach and associated modified version of the Rand index for weakly-labeled data is introduced to address these issues. Results are evaluated with the new index and compared to traditional segmentation evaluation methods. Experimental results on a SAS data set containing must-link and cannot-link labels show that our Weakly-Labeled Rand index scores segmentations appropriately in reference to qualitative performance and is more suitable than traditional quantitative metrics for scoring weakly-labeled data.