Trusting Semantic Segmentation Networks

TL;DR

This paper analyzes failure modes of semantic segmentation models and evaluates how uncertainty metrics like entropy can predict misclassifications, enhancing trust in model outputs.

Contribution

It provides the first systematic analysis of segmentation failure prediction using uncertainty metrics across multiple models and datasets.

Findings

Entropy correlates strongly with misclassification.

Simple uncertainty measures can effectively predict errors.

High recall rates achieved in identifying failures.

Abstract

Semantic segmentation has become an important task in computer vision with the growth of self-driving cars, medical image segmentation, etc. Although current models provide excellent results, they are still far from perfect and while there has been significant work in trying to improve the performance, both with respect to accuracy and speed of segmentation, there has been little work which analyses the failure cases of such systems. In this work, we aim to provide an analysis of how segmentation fails across different models and consider the question of whether these can be predicted reasonably at test time. To do so, we explore existing uncertainty-based metrics and see how well they correlate with misclassifications, allowing us to define the degree of trust we put in the output of our prediction models. Through several experiments on three different models across three datasets, we show that simple measures such as entropy can be used to capture misclassification with high recall rates.