Interactive Learning for Semantic Segmentation in Earth Observation

TL;DR

This paper introduces DISCA, an interactive learning framework that refines semantic segmentation maps in Earth observation images by continually adapting neural networks with sparse user annotations, improving accuracy especially in domain adaptation scenarios.

Contribution

The paper presents a novel interactive learning method for semantic segmentation that enhances accuracy through continual adaptation with minimal user input.

Findings

Achieves up to 4.7% IoU improvement with ten user clicks.

Effective in domain adaptation scenarios.

Demonstrates benefits on three different datasets.

Abstract

Dense pixel-wise classification maps output by deep neural networks are of extreme importance for scene understanding. However, these maps are often partially inaccurate due to a variety of possible factors. Therefore, we propose to interactively refine them within a framework named DISCA (Deep Image Segmentation with Continual Adaptation). It consists of continually adapting a neural network to a target image using an interactive learning process with sparse user annotations as ground-truth. We show through experiments on three datasets using synthesized annotations the benefits of the approach, reaching an IoU improvement up to 4.7% for ten sampled clicks. Finally, we exhibit that our approach can be particularly rewarding when it is faced to additional issues such as domain adaptation.