Deep Probability Segmentation: Are segmentation models probability estimators?

TL;DR

This paper investigates whether deep segmentation models can inherently serve as reliable probability estimators, emphasizing the importance of uncertainty quantification in pixel-wise classification tasks.

Contribution

The study evaluates the impact of Calibrated Probability Estimation (CaPE) on segmentation models, highlighting their potential as effective probabilistic estimators with inherent calibration capabilities.

Findings

CaPE improves calibration but less so than in classification tasks

Segmentation models inherently provide better probability estimates

Dataset size and bin optimization influence calibration effectiveness

Abstract

Deep learning has revolutionized various fields by enabling highly accurate predictions and estimates. One important application is probabilistic prediction, where models estimate the probability of events rather than deterministic outcomes. This approach is particularly relevant and, therefore, still unexplored for segmentation tasks where each pixel in an image needs to be classified. Conventional models often overlook the probabilistic nature of labels, but accurate uncertainty estimation is crucial for improving the reliability and applicability of models. In this study, we applied Calibrated Probability Estimation (CaPE) to segmentation tasks to evaluate its impact on model calibration. Our results indicate that while CaPE improves calibration, its effect is less pronounced compared to classification tasks, suggesting that segmentation models can inherently provide better probability estimates. We also investigated the influence of dataset size and bin optimization on the effectiveness of calibration. Our results emphasize the expressive power of segmentation models as probability estimators and incorporate probabilistic reasoning, which is crucial for applications requiring precise uncertainty quantification.