UniRSCD: A Unified Novel Architectural Paradigm for Remote Sensing Change Detection

2025-11-22

AI Summary

Existing remote sensing change detection (RSCD) methods rely on expert-designed, task-specific decoders to compensate for encoding information loss, resulting in high model selection uncertainty and poor cross-task generalization. To address this, we propose UniRSCD—a unified framework built upon a state-space model backbone. We introduce the first frequency-aware change prompt generator as a unified encoder that dynamically fuses high- and low-frequency features to construct a task-adaptive shared representation space, eliminating the need for task-customized decoders. Furthermore, hierarchical feature interaction and task-adaptive output mapping enable end-to-end unified modeling across multi-granularity tasks—including binary change detection, semantic change detection, and building damage assessment. UniRSCD achieves state-of-the-art performance on five benchmark datasets (LEVIR-CD, SECOND, xBD, WHU-CD, and DSIFN-CD), significantly enhancing both generality and practicality of RSCD methods.

Technology Category

• Reasoning under Uncertainty: Uncertainty Representations
• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

In recent years, remote sensing change detection has garnered significant attention due to its critical role in resource monitoring and disaster assessment. Change detection tasks exist with different output granularities such as BCD, SCD, and BDA. However, existing methods require substantial expert knowledge to design specialized decoders that compensate for information loss during encoding across different tasks. This not only introduces uncertainty into the process of selecting optimal models for abrupt change scenarios (such as disaster outbreaks) but also limits the universality of these architectures. To address these challenges, this paper proposes a unified, general change detection framework named UniRSCD. Building upon a state space model backbone, we introduce a frequency change prompt generator as a unified encoder. The encoder dynamically scans bitemporal global context information while integrating high-frequency details with low-frequency holistic information, thereby eliminating the need for specialized decoders for feature compensation. Subsequently, the unified decoder and prediction head establish a shared representation space through hierarchical feature interaction and task-adaptive output mapping. This integrating various tasks such as binary change detection and semantic change detection into a unified architecture, thereby accommodating the differing output granularity requirements of distinct change detection tasks. Experimental results demonstrate that the proposed architecture can adapt to multiple change detection tasks and achieves leading performance on five datasets, including the binary change dataset LEVIR-CD, the semantic change dataset SECOND, and the building damage assessment dataset xBD.

Problem

• Unified architecture eliminates specialized decoders for different change detection tasks
• Dynamic encoder integrates high-frequency details with low-frequency global information
• Framework accommodates varying output granularities across multiple remote sensing datasets

Innovation

• State space model backbone for global context scanning
• Frequency change prompt generator integrating high-low frequency details
• Unified decoder with hierarchical interaction and task-adaptive mapping