Uncertainty and Energy based Loss Guided Semi-Supervised Semantic Segmentation

TL;DR

This paper introduces a semi-supervised semantic segmentation method that leverages aleatoric uncertainty and energy-based loss to improve training with pseudo labels, demonstrating enhanced performance over existing approaches.

Contribution

It proposes a novel intersection-union pseudo supervised network utilizing uncertainty and energy modeling, advancing semi-supervised segmentation techniques.

Findings

Improved segmentation accuracy compared to state-of-the-art methods.

Effective modeling of data uncertainty through dual predictive branches.

Enhanced learning via energy-based loss in pseudo supervision.

Abstract

Semi-supervised (SS) semantic segmentation exploits both labeled and unlabeled images to overcome tedious and costly pixel-level annotation problems. Pseudolabel supervision is one of the core approaches of training networks with both pseudo labels and ground-truth labels. This work uses aleatoric or data uncertainty and energy based modeling in intersection-union pseudo supervised network.The aleatoric uncertainty is modeling the inherent noise variations of the data in a network with two predictive branches. The per-pixel variance parameter obtained from the network gives a quantitative idea about the data uncertainty. Moreover, energy-based loss realizes the potential of generative modeling on the downstream SS segmentation task. The aleatoric and energy loss are applied in conjunction with pseudo-intersection labels, pseudo-union labels, and ground-truth on the respective network branch. The comparative analysis with state-of-the-art methods has shown improvement in performance metrics.