Detect Changes Like Humans: Incorporating Semantic Priors for Improved Change Detection

Yuhang Gan, Wenjie Xuan, Zhiming Luo, Lei Fang, Zengmao Wang, Juhua Liu, Bo Du

Abstract: When given two similar images, humans identify their differences by comparing the appearance (e.g., color and texture) with the help of semantics (e.g., objects and relations). However, mainstream binary change detection models adopt a supervised training paradigm, where the annotated binary change map is the main constraint. Thus, such methods primarily emphasize difference-aware features between bitemporal images, and the semantic understanding of changed landscapes is undermined, resulting in limited accuracy in the face of noise and illumination variations. To this end, this article explores incorporating semantic priors from visual foundation models to improve the ability to detect changes. First, we propose a semantic-aware change detection network (SA-CDNet), which transfers the knowledge of visual foundation models (i.e., FastSAM) to change detection. Inspired by the human visual paradigm, a novel dual-stream feature decoder is derived to distinguish changes by combining semantic-aware features and difference-aware features. Second, we explore a single-temporal pretraining strategy for better adaptation of visual foundation models. With pseudo-change data constructed from single-temporal segmentation datasets, we employ an extra branch of the proxy semantic segmentation task for pretraining. We explore various settings like dataset combinations and landscape types, thus providing valuable insights. Experimental results on five challenging benchmarks demonstrate the superiority of our method over the existing state-of-the-art methods. The code is available at https://github.com/DREAMXFAR/SA-CDNet

External IDs:doi:10.1109/tgrs.2025.3608816