Style-Transfer-Based Unsupervised Change Detection From Heterogeneous Images

Heterogeneous images are captured through different wavelength bands, providing rich and complementary information for change detection (CD), and domain transformation has emerged as a popular and effective solution. However, existing domain-transformation-based CD methods overly rely on the quality of reconstructed features, making them appear inadequate for practical complex scenarios. In this article, we propose a style-transfer-based CD (STCD) method through unsupervised learning. STCD improves the quality and enhances the robustness of the reconstructed images by simultaneously employing a cautious labeling strategy and classification. Specifically, we initially convert the two heterogeneous images provided into a shared domain by constructing a convolutional autoencoder based on adaptive instance normalization, which improves the quality of reconstructed features and mitigates data heterogeneity. Furthermore, we extract some significant pixel pairs based on fuzzy local information $c$-means to reduce the overreliance on reconstructed features. Then, we propose a dynamic superpixel-based label assignment rule to increase the reliable pseudo-labels employed in training a binary classifier. Finally, STCD achieves great CD results even with poor reconstruction quality. Experimental results conducted on four heterogeneous datasets demonstrate the effectiveness of STCD over other related CD methods.