SiamSeg: Self-Training with Contrastive Learning for Unsupervised Domain Adaptation Semantic Segmentation in Remote Sensing

TL;DR

SiamSeg introduces a self-training method with contrastive learning for unsupervised domain adaptation in remote sensing image segmentation, significantly improving performance across diverse datasets by enhancing semantic feature learning.

Contribution

This paper presents a novel integration of contrastive learning into self-training for UDA in RS segmentation, addressing domain shift challenges and achieving state-of-the-art results.

Findings

Outperforms existing UDA methods on RS datasets

Achieves state-of-the-art segmentation accuracy

Enhances semantic feature representation in target domain

Abstract

Semantic segmentation of remote sensing (RS) images is a challenging yet essential task with broad applications. While deep learning, particularly supervised learning with large-scale labeled datasets, has significantly advanced this field, the acquisition of high-quality labeled data remains costly and time-intensive. Unsupervised domain adaptation (UDA) provides a promising alternative by enabling models to learn from unlabeled target domain data while leveraging labeled source domain data. Recent self-training (ST) approaches employing pseudo-label generation have shown potential in mitigating domain discrepancies. However, the application of ST to RS image segmentation remains underexplored. Factors such as variations in ground sampling distance, imaging equipment, and geographic diversity exacerbate domain shifts, limiting model performance across domains. In that case, existing ST methods, due to significant domain shifts in cross-domain RS images, often underperform. To address these challenges, we propose integrating contrastive learning into UDA, enhancing the model's ability to capture semantic information in the target domain by maximizing the similarity between augmented views of the same image. This additional supervision improves the model's representational capacity and segmentation performance in the target domain. Extensive experiments conducted on RS datasets, including Potsdam, Vaihingen, and LoveDA, demonstrate that our method, SimSeg, outperforms existing approaches, achieving state-of-the-art results. Visualization and quantitative analyses further validate SimSeg's superior ability to learn from the target domain. The code is publicly available at https://github.com/woldier/SiamSeg.